(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Children's Library | Biodiversity Heritage Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "Structure/property relationships in methacrylate/dimethacrylate polymers for dental applications"

STRUCTURE/PROPERTY RELATIONSHIPS IN 
METHACRYLATE/DIMETHACRYLATE POLYMERS FOR DENTAL 

APPLICATIONS 



By 
JEREMY JOHN MEHLEM 



A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL 

OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT 

OF THE REQUIREMENTS FOR THE DEGREE OF 

DOCTOR OF PHILOSOPHY 

UNIVERSITY OF FLORIDA 

2001 



[This dissertation is dedicated to all of those who gave their love and support during the 
writing of the dissertation, especially my wife Kacy.] 









ACKNOWLEDGMENTS 

I would like to thank my advisor and doctoral committee, Dr. Anthony Brennan, 
without whose enthusiasm, knowledge, and guidance none of this would be possible. I 
would also like to thank the members of the supervisory committee for their valuable 
input: Dr. Elliot Douglas, Dr. Christopher Batich, Dr. Ronald Baney, and Dr. Kenneth 
Wagener. 

I cannot thank enough all of my colleagues who gave the support and 
collaboration necessary to make it through this onerous process: Mark Schwarz, Jennifer 
Russo, Jeanne Macdonald, Dr. Luxsamee Plangsangmas, Xiaomei Qian, Derek Lincoln, 
Arthur Gavrin, Jamie Rhodes, Lee Zhao, Dr. Brent Gila, Dr. Drew Amery, Bob Hadba, 
Dr. Chris Widenhouse, Dr. Rodrigo Orifice, Dr. James Marrotta, and Paul Martin. I 
would also like to thank the next generation of colleagues for bringing new ideas and 
perspectives to our discussions: Clayton Bohn, Adam Feinberg, Wade Wilkerson, Nikhil 
Kothurkar, Amy Gibson, Charles Seegert, Leslie Wilson, Brian Hatcher, Brendan Hauser, 
Bryan Cuevas, Josh Stopek, and all the other students, faculty, and staff that made my 
experience at the University of Florida memorable. 

I cannot express enough gratitude to Dr. Jesse Arnold, Dr. Tom Miller and Dr. 
Mike Zamora for their guidance early in my career as a graduate student and their 
continued advice as I go into the workplace. 

Special thanks goes out to Ms. Tonya Brevaldi for her help with NMR 
spectroscopy and Mrs. Jeanne Macdonald for her help with DSC measurements. I would 

iii 



also like to thank Dr. C. Russell Bowers and his graduate student, Tony Zook, for all of 
their help making l29 xenon NMR measurements. I could not have completed the work 
without their collaboration. 

I would be remiss not to recognize the aid that the program assistants provide. I 
would also like to thank Dr. Peter Ifju for his help in making modulus measurement on 
our standards and his impromptu mentoring at our board meetings. 

I must acknowledge the Center for the Development of Alternatives to Dental 
Amalgam at the University of Florida, especially Mr. Ben Lee in the dental biomaterials 
laboratory and the NIH-NIDR (grant no. 2-P50-DE 09307) for funding my research and 
allowing me the opportunity to attend and present my work at national conferences. 
I would like to thank my parents and family for supporting my continued education. 
Last, but not least, I would like to thank my wife, Kacy, for her love and support through 
the last years of my education. 






IV 



TABLE OF CONTENTS 

page 

ACKNOWLEDGMENTS iii 

LIST OF TABLES vii 

LIST OF FIGURES x 

ABSTRACT xiv 

1 INTRODUCTION 1 

2 BACKGROUND 5 

History of Dental Restorative Materials 5 

Components of Dental Composites 6 

Areas for Improvement 8 

Network Structure 8 

Bis-GMA Analogs and Modifications 12 

Monomethacrylates 16 

Polymerization Shrinkage 18 

Anhydrides in Dental Materials 25 

Water Absorption 28 

Mechanical Properties 29 

1 7Q 

Xenon Nuclear Magnetic Resonance Spectroscopy 32 

Dynamic Mechanical Spectroscopy 42 

Summary 44 

3 ALTERNATE DILUENT SYSTEMS FOR BIS-GMA AND BIG-GMA 

ANALOGS 47 

Relevant Background 47 

Materials and Methods 52 

Results and Discussion 60 

Physical Properties 60 

Flexure Testing 69 

Fracture Toughness and Tensile Testing 76 

Dynamic Mechanical Spectroscopy 83 



Comparison of Alternate Comonomer Systems by Weight to Triethylene Glycol 

Dimethacrylate 110 

Flexure Testing 1 10 

Tensile Strength and Fracture Toughness 115 

Conclusions and Future Work 120 



4 CHARACTERIZING THE HETEROGENITY OF DIMETHACRYLATE 
POLYMERS 123 

Relevant Background 123 

Materials and Methods 126 

Results and Discussion 131 

Atomic Force Microscopy 131 

Xenon-129NMR 145 

Conclusions and Future Work 156 



5 EVALUATION OF THE EFFICIENCY OF THE INCORPORATION OF NADIC 
METHYL ANHYDRIDE, NORBORNENE BASED COMPOUNDS AND MALEIC 
ANHYDRIDE INTO METHACRYLATE-BASED DENTAL RESINS 158 

Relevant Background 158 

Materials and Methods 161 

Results and Discussion 163 

Conclusions 177 



6 CLOSING REMARKS 181 

LIST OF REFERENCES 185 

BIOGRAPHICAL SKETCH 196 



VI 



LIST OF TABLES 



Table Page 

Table 2.1: Volumetric shrinkage values and modulus values for Bis-GMA molecules 
esterified with various-length aliphatic acids. The number next to C 
represents the number of carbon atoms in the aliphatic chain 21 

Table 2.2: Mechanical properties and water sorption of various unfilled dimethacrylate 

polymers wet and dry 30 

Table 2.3: Degree of conversion values for Bis-GMA/TEGDMA polymers systems 

obtained by Ferracane 32 

Table 2.4: Glass transition temperature estimates obtained by Wilson and Turner 44 

Table 3.1: Materials used 53 

Table 3.2: Molar fractions, weight fractions, and viscosities of Bis-MEPP/CHMA, Bis- 

MEPP/t-BCHMA, and Bis-MEPP/PEMA resins 55 

Table 3.3: Predicted theoretical and measured shrinkage values for Bis-MEPP-based 

copolymer systems 64 

Table 3.4: Physical properties of Bis-MEPP/PEMA systems 65 

Table 3.5: Physical properties of Bis-MEPP/CHMA systems 66 

Table 3.6: Physical properties of Bis-MEPP/t-BCHMA systems 67 

Table 3.7: Physical properties of Bis-MEPP/EGDMA-type resins 67 

Table 3.8: Degree of conversion values based on percent vinyl bonds remaining in Bis- 
MEPP/PEMA, Bis-MEPP/CHMA, and Bis-MEPP/t-BCHMA systems 69 

Table 3.9: Degree of conversion values based on percent vinyl bonds remaining in Bis- 
MEPP/EDGMA-type systems, and the Bis-MEPP/PEMA, Bis-MEPP/CHMA, 
and Bis-MEPP/t-BCHMA systems formulated to approximately 35 weight 
percent diluent 69 



vn 



Table 3.10: Pair-wise comparison of the modulus values of different polymer systems by 

mol.% formulation in the wet state using Student's t-test 73 

Table 3.11: Student's t-tests comparing tensile strengths of the various Bis-MEPP/t- 

BCHMA formulations in the wet and dry state and the same formulation in its 
wet and dry state 79 

Table 3.12: Student's t-tests comparing wet fracture toughness values of the various 

formulations in the wet state 82 

Table 3.13: Statistics on tan 8 plots and glass transition activations energies for polymers 

in the dry state at 1 Hertz 88 

Table 3.14: WLF parameters used to obtain the master curve fits for the Bis- 

MEPP/PEMA and Bis-MEPP/CHMA systems formulated to a glass transition 
temperature of 135°C 109 

Table 4. 1 : Feedback parameters for Tapping Mode™ Atomic Force Microscopy 127 

Table 4.2: Parameters used to collect standard xenon spectra of Bis-MEPP(XDT) 

samples 128 

Table 4.3: Hahn echo sequence parameters used for water self-diffusion coefficient 

measurements 129 

Table 4.4: Parameters used in pulse sequence to measure the self-diffusion coefficient of 

xenon in Bis-MEPP(XDT) 131 

Table 4.5: Xenon NMR spectrum parameters for Bis-MEPP(XDT) at various stages of 

cure 148 

Table 4.6: Degree of conversion values, diffusion coefficients, and domain sizes of Bis- 

MEPP(XDT) at various stages of cure 155 

Table 5.1: 2-phenylethyl methacrylate/nadic methyl anhydride/maleic anhydride 

monomer compositions in mol.% 162 

Table 5.2: FTIR and DSC results from 2-phenylethyl methacrylate/nadic methyl 

anhydride based model compounds 167 

Table 5.3: Molar mass averages from GPC for 2-phenylethyl methacrylate-nadic methyl 

anhydride copolymers 167 

Table 5.4: FTIR data and DSC from 2-phenylethyl methacrylate/nadic methyl 

anhydride/maleic anhydride based model compounds 171 

Table 5.5: Molar mass averages from GPC for 2-phenylethyl methacrylate-nadic methyl 

anhydride-maleic anhydride copolymers 171 






LIST OF FIGURES 



Figure Page 

Figure 2.1: The structure of 2,2-bis(4-(2-hydroxy-3-methacryloyloxyprop-l- 

oxy)phenol)propane (Bis-GMA) 15 

Figure 2.2: Dimethacrylates synthesized from various diphenols 15 

Figure 2.3: Bis-GMA analogs referred to in Table 2.2 33 

Figure 2.4: Ethoxylated Bis-GMA molecules 34 

Figure 2.5: 2,2'-bis-(4-methacryloylpropoxyphenyl) propane (Bis-MPPP) 34 

Figure 3.1: The structure of 2-phenyloxyethyl methacrylate (PEMA) 50 

Figure 3.2: The structure of cyclohexyl methacrylate (CHMA) 50 

Figure 3.3: The structure of t-butylcyclohexyl methacrylate (t-BCHMA) 50 

Figure 3.4: The basic structure of poly (ethylene glycol dimethacrylate) analogs 54 

Figure 3.5: Schematic of mold used for sample manufacture. Source:[99] 57 

Figure 3.6: Viscosity measurements on Bis-MEPP/PEMA, Bis-MEPP/t-BCHMA, and 

Bis-MEPP/CHMA resins at 25°C 62 

Figure 3.7: Viscosity measurements of various diluents in the Bis-MEPP-based resin 

systems diluted at approximately 35 weight percent 63 

Figure 3.8: Flexure modulus values of wet and dry Bis-MEPP/CHMA, Bis- 
MEPP/PEMA, and Bis-MEPP/t-BCHMA polymers at 37°C. The legend 
represents the monomers used to diluent the Bis-MEPP monomer 71 

Figure 3.9: Tensile strength values of wet and dry Bis-MEPP/PEMA, Bis- 
MEPP/CHMA, and Bis-MEPP/t-BCHMA polymers at 37°C 84 

Figure 3.10: Fracture toughness values of wet and dry Bis-MEPP/PEMA, Bis- 
MEPP/CHMA, and Bis-MEPP/t-BCHMA polymers at 37°C 85 



Figure 3.1 1: Log E' and tan 5 response of dry Bis-MEPP/PEMA polymers at 1 hertz 92 

Figure 3.12: Log E' and tan 8 response of dry 70/30 Bis-MEPP/PEMA polymers at 

various frequencies 93 

Figure 3.13: Log E' and tan 8 response of wet Bis-MEPP/PEMA polymers at 1 hertz 94 

Figure 3.14: Log E' and tan 8 response of dry Bis-MEPP/CHMA polymers at 1 hertz 97 

Figure 3.15: Log E' and tan 8 response of wet Bis-MEPP/CHMA polymers at 1 hertz 98 

Figure 3.15: Log E' and tan 8 response of wet Bis-MEPP/CHMA polymers at 1 hertz 98 

Figure 3.16: Log E' and tan 8 response of dry 70/30 Bis-MEPP/CHMA polymers at 

various frequencies 99 

Figure 3.17: Log E' and tan 8 response of dry Bis-MEPP/t-BCHMA polymers at 1 hertz. 101 

Figure 3.18: Log E' and tan 8 response of wet Bis-MEPP/t-BCHMA polymers at 1 hertz. 102 

Figure 3.19: Log E' and tan 8 response of dry Bis-MEPP/EDGMA type polymers at 1 

hertz 104 

Figure 3.20: Log E' and tan 8 response of wet Bis-MEPP/EDGMA type polymers at 1 

hertz 105 

Figure 3.21: Log E' and tan 8 response of dry Bis-MEPP/PEMA, Bis-MEPP/CHMA, 

and Bis-MEPP/t-BCHMA polymers formulated to a Tg of 135°C 106 

Figure 3.22: Master curve plots of the Bis-MEPP/PEMA (top) and Bis-MEPP/CHMA 
(bottom) formulated to glass transition temperatures of 135°C (the frequency 
range of the x-axis is le-16 to le32 s" 1 .) 11 1 

Figure 3.23: The plot of shift factors (a t ) calculated from the WLF equation for the Bis- 
MEPP/PEMA and Bis-MEPP/CHMA systems formulated to a Tg of 135°C 1 12 

Figure 3.24: Log E' and tan 8 response of dry Bis-MEPP/PEMA with normal cure and 

post cure at 140°C Bis-MEPP/CHMA 113 

Figure 3.25: Flexure modulus values of wet and dry 65/35 weight percent Bis- 
MEPP/TEGDMA, Bis-MEPP/EGDMA multicomponent type, Bis- 
MEPP/CHMA, Bis-MEPP/PEMA, and Bis-MEPP/t-BCHMA polymers at 
37°C. The legend represents the monomers used to dilute the Bis-MEPP 
monomer 116 

Figure 3.26: Tensile strength values of wet and dry 65/35 weight percent Bis- 
MEPP/TEGDMA, Bis-MEPP/EGDMA multicomponent type, Bis- 
MEPP/CHMA, Bis-MEPP/PEMA, and Bis-MEPP/t-BCHMA polymers at 



XI 



37°C. The legend represents the monomers used to dilute the Bis-MEPP 
monomer 118 

Figure 3.27: Fracture toughness values of wet and dry 65/35 weight percent Bis- 
MEPP/TEGDMA, Bis-MEPP/EGDMA multicomponent type, Bis- 
MEPP/CHMA, Bis-MEPP /PEMA, and Bis-MEPP/t-BCHMA polymers at 
37°C. The legend represents the monomers used to dilute the Bis-MEPP 
monomer 119 

Figure 4. 1 : Pulse schemes of Hahn echo pulsed field gradient and stimulated echo pulsed 

field gradient sequence 130 

Figure 4.2: Phase and topographic AFM images of a fracture surface of a 60/40 Bis- 

MEPP/TEGDMA polymer at 500 nm scale 136 

Figure 4.3: Phase and topographic AFM images of a fracture surface of a 60/40 Bis- 

MEPP/TEGDMA polymer at a 1 urn scale 137 

Figure 4.4: Phase and topographic AFM images of a microtomed surface of a 70/30 Bis- 

MEPP/PEMA polymer at 500 nm scale 139 

Figure 4.5: Phase and topographic AFM images of a microtomed surface of a 70/30 Bis- 

MEPP/PEMA polymer at 1 um scale 140 

Figure 4.6: Phase and topographic AFM images of a microtomed surface of a 60/40 Bis- 

MEPP/TEGDMA polymer immersed in acetone and then dried 143 

Figure 4.7: The 3-D perspective of an AFM phase image of a microtomed surface of a 

60/40 Bis-MEPP/TEGDMA polymer immersed in acetone and then dried 144 

Figure 4.8: Phase and topographic AFM images of a microtomed surface of a 70/30 Bis- 
MEPP/PEMA polymer imaged with a set-point value of 0.40 volts at 500 nm 
scale 146 

Figure 4.9: Phase and topographic AFM images of a microtomed surface of a 70/30 Bis- 
MEPP/PEMA polymer post cured for 2 hours at 140°C immediately after 
light-curing imaged with a set-point value of 0.40 volts at 500 nm scale 147 

Figure 4.10: Xenon-129 spectra of Bis-MEPP(XDT) monomer at 25°C 149 

Figure 4.1 1: Xenon-129 spectra at 25°C of Bis-MEPP(XDT) cured for 30 seconds at 

25°C 151 

Figure 4.12: Xenon-129 spectra at 25°C of Bis-MEPP(XDT) cured for 5 minutes 153 

Figure 5.1: The structure of a-methylene-y-butyrolactone 159 



XII 



Figure 5.2: The structure of nadic methyl anhydride (NMA)(methyl-5-norbornene-2,3 

dicarboxylic acid anhydride) 159 

Figure 5.3: The structure of maleic anhydride (MA) 159 

Figure 5.4: The structure of 2-phenylethyl methacrylate 164 

Figure 5.5: The structure of 5-norbornene-2,3 dicarboxylic acid anhydride 164 

Figure 5.6: The structure of 5-norbornene-2-carboxaldehyde 165 

Figure 5.7: The structures of 5-norbornene-2-butane and 5-norbornene-2-hexane 165 

Figure 5.8: FTIR spectra of PMA-co-40%NMA, PMA, poly (PMA-co-40%NMA), and 

poly(PMA-co-7.5%NMA-co-22.5%MA) 168 

Figure 5.9: The proton NMR spectra of 2-phenylethyl methacrylate in deuterated DMF 

referenced toTMS 173 

Figure 5.10: The proton NMR spectra of nadic methyl anhydride in deuterated DMF 

referenced to TMS 175 

Figure 5.1 1: The proton NMR spectra of maleic anhydride in DMF referenced to TMS.... 176 

Figure 5.12: The proton NMR spectra of poly (PMA-co-40%NMA) in deuterated DMF 

referenced to TMS 178 

Figure 5.13: The proton NMR spectra of poly (PMA-co-7.5%NMA-co-22.5%MA) in 

deuterated DMF referenced to TMS 179 

Figure 5.14: The proton NMR spectra of deuterated DMF referenced to TMS 180 












xm 



Abstract of Dissertation Presented to the Graduate School 

of the University of Florida in Partial Fulfillment of the 

Requirements for the Degree of Doctor of Philosophy 

STRUCTURE/PROPERTY RELATIONSHIPS IN 
METHACRYLATE/DIMETHACRYLATE POLYMERS FOR DENTAL 

APPLICATIONS 



By 

Jeremy John Mehlem 

May, 2001 

Chair: Anthony B. Brennan 

Major Department: Materials Science and Engineering 

Since its invention Bis-GMA or one of its analogs has been the main component 

of the polymer portion of composites for dental restorations. The need for dilution of 

Bis-GMA and its analogs to optimize its properties has long been recognized. Bis-GMA 

is a highly viscous monomer. This high viscosity leads to early vitrification, which limits 

conversion during cure. This viscosity also limits filler loading. Vitrification at low 

conversions leads to heterogeneous systems composed of low and high cross-link density 

phases. The low cross-link density phases behave as defects in the system; therefore, if 

the amount of low cross-link density phases in the system can be reduced and a more 

uniform network structure can be achieved, then the mechanical properties of the resin 

can be improved. Since the increase in viscosity during cure causes vitrification, it is 

logical that a system with a low initial viscosity will delay the onset of vitrification. 

Reactive diluents such as triethylene glycol dimethacrylate (TEGDMA) are effective at 

xiv 



lower levels. However, large amounts negatively affect matrix properties by increasing 
polymerization shrinkage and water sorption. Shrinkage has been cited as one of the 
main deficiencies in dental composites. The goal of this project is to improve upon 
standard viscosity modifying comonomers such as triethylene glycol dimethacrylate. The 
comonomers that were explored were phenyloxyethyl methacrylate, cyclohexyl 
methacrylate, and tert-butylcylcohexyl methacrylate. Multicomponent systems based on 
analogs of ethylene glycol dimethacrylates with different length ethyl glycol chains were 
also examined. The substitution of monomethacrylates for TEGDMA as a comonomer 
resulted in enhanced or negligible affects on the mechanical properties of Bis-MEPP 
based polymer systems while reducing polymerization shrinkage. 

I29 Xenon NMR and TappingMode™ AFM were used to characterize the 
heterogeneity of dimethacrylates systems during their cure cycle as well as in their final 
state. Using these methods the size of the high and low cross-link density phase was 
examined and determined to be on the order of 50-150 nanometers. 

Model compounds based on phenylethyl methacrylate were formulated to 
determine how of nadic methyl anhydride and maleic anhydride incorporate into 
dimethacrylate resin systems. 



xv 



CHAPTER 1 
INTRODUCTION 

Since the discovery of the toxicity of mercury vapor, even in small amounts, there 

has been a large push in the dental industry for alternatives to dental amalgam materials. 

This need led to the introduction of polymer-based composites as restorative dental 

materials. Most polymer-based composites are composed of dimethacrylate polymers and 

inorganic fillers. Polymer-based composites have many redeeming qualities such as ease 

of use, high aesthetic appeal, and, of course, a less toxic alternative to mercury-based 

amalgams. Polymer-based composites are not without their drawbacks. The shrinkage 

that occurs upon polymerization is still cited as one of the major reasons for the failure of 

dental composites in the oral environment. This shrinkage results in stresses that cause 

defects with the composite matrix and debonding from the tooth restoration interface, 

resulting in poor marginal adaptation, marginal leakage, and subsequent recurrent decay. 

Most modern restorative materials are visible light-cured composites based on high molar 

mass monomers such as 2,2'-bis(4-(2-hydroxy-3-methacryloyloxyprop-l-oxy) 

phenyl)propane (Bis-GMA), in reference to its precursor materials bis-phenol A and 

glycidyl methacrylate, 2,2-bis-(4-methacryloylethoxyphenyl) propane (Bis-MEPP), 

sometimes referred to as ethoxylated Bis-GMA, and urethane dimethacrylates. The need 

for dilution of Bis-GMA and its analogs to optimize its properties has long been 

recognized. Bis-GMA is a highly viscous monomer (2980 Pa«s [1]), which leads to early 

vitrification, limited conversion during cure, and limited filler loading. Vitrification is 

defined as the point where the glass transition temperature of the reacting system reaches 

1 



the reaction temperature. Polymerization becomes diffusion limited at the vitrification 
point. Vitrification at low conversions leads to heterogeneous systems composed of low 
molar mass phases and highly cross-linked phases. The lower molar mass or low cross- 
link density phases behave as defects in the system. The mechanical properties of the 
polymer can be improved, if the volume fraction of the low molar mass or the low cross- 
link density material of the system can be reduced. IF the increase in viscosity during 
cure causes vitrification, it is logical that a system with a low initial viscosity will delay 
the onset of vitrification. This hypothesis is supported by the relationship between initial 
viscosity and degree of conversion[2-4]. Reactive diluents such as triethylene glycol 
dimethacrylate (TEGDMA) are effective at lower levels, but large amounts negatively 
affect matrix properties by increasing polymerization shrinkage and water sorption. The 
polymerization shrinkage of poly (TEGDMA) is 1 1.9 % and it water uptake is 6.0 weight 
percent. The polymerization shrinkage and water uptake of poly (Bis-GMA) are 6.4 % 
and 3.05 weight percent, respectively[5]. Excessive polymerization shrinkage results in 
tensile stresses in matrix materials adjacent to the glass filler. These stresses lower 
mechanical properties, i.e., modulus, stress at break. However, the lower viscosity 
achieved with the addition of reactive diluents also allows higher glass loadings, which 
improves mechanical properties and reduces the shrinkage in the overall system. 

From this point in this work the term "light-cured" will be used to refer to light- 
initiated polymerization of the vinyl bonds. In light-cured systems the polymerization is 
initiated by radicals that are formed upon exposure to light and subsequently react with 
the vinyl bonds. The first light-cured systems were based on compounds that were 
activated in the UV region. More modern systems use compounds that are activated in 






the blue region. The actual polymerization process is similar to a chemically initiated 
polymerization except for one very important aspect, i.e, light-cured systems do not 
require mixing, which is very significant clinically. 

To date the three primary approaches for reducing polymerization shrinkage have 
been ring opening polymerizations, swellable clays, and monomers with higher specific 
volumes. All of these methods have their drawbacks [6-8]. 

Many dimethacrylates form heterogeneous structures that are characterized by 
areas of highly cross-linked polymer, which surround or contain areas of low cross-link 
density. This is a result of the evolution of structure during cure. Kinetic gelation 
modeling indicates that this heterogeneity has its roots early in the polymerization 
process and manifests itself in the form of isolated branched molecules with large 
numbers of pendant double bonds [9]. Modeling based on percolation theory also 
predicts the formation of these molecules [10]. These molecules, sometimes referred to 
as microgels, have been isolated and detected with dynamic light scattering [11]. 
Polymerization continues preferentially within and around these microgels. These 
molecules ultimately grow together resulting in the trapping of radicals and unreacted 
material. As a polymer network forms, the onset of vitrification limits monomer 
diffusion, the mobility of pendant groups, and the degree of conversion. The pendant 
groups and monomers act as defects. These structures along with the limited amount of 
active chains in the network restrict the mechanical and physical properties of these 
materials. 

The goal of this work is to examine viscosity modifying comonomer systems 
based on monomethacrylates, or multicomponent analogs of ethylene glycol 



dimethacrylate that will reduce the polymerization shrinkage and improve or negligibly 
affect mechanical properties of Bis-MEPP based polymers when compared to the 
standard comonomer TEGDMA. The specific aims of this work are as follows: Aim 1: 
To test the hypothesis that monofunctional methacrylates with a cyclic or aromatic 
pendent group can reduce polymerization shrinkage while improving or negligibly 
affecting mechanical properties. Aim 2: To test the hypothesis that heterogeneity can be 
reduced in polymers designed with the same glass transition temperature by having a 
lower viscosity monomer system. Aim 3: To test the hypothesis that multicomponent 
diluent systems based on analogs of triethylene glycol dimethacrylate with a range of 
molar masses can lower viscosity more effectively than triethylene glycol dimethacrylate. 
The multicomponent system will maintain the same weight average molar mass as 
triethylene glycol dimethacrylate. Aim 4: To test the hypothesis that norbornene 
monomers that react to form high glass transition temperature polymers can be 
incorporated into dental polymers. Aim 5: To test the hypothesis that l29 xenon NMR 
and TappingMode™ Atomic Force Microscopy with phase mapping can be used to 
characterize the size and distribution of heterogeneity in dimethacrylate polymers and 
how it changes with conversion. 



CHAPTER 2 
BACKGROUND 

History of Dental Restorative Materials 

The direct placement of filling material is a common dental procedure. Polymer- 
ceramic composite restorations are a favorite with patients for aesthetic reasons as well as 
ease of handling [12]. The shortcomings of the original polymer-based restorations have 
driven the evolution of dental materials to their current state. 

The first polymer-based fillings were acrylates [13]. They were used from about 
1945 until 1960, but were discarded because of their inadequate mechanical properties, 
large dimensional changes with temperature, large volumetric shrinkage during curing, 
and low stain resistance. 

In 1962 Bowen developed the first Bis-GMA- based composite system (Figure 
2.1). His work resulted in a molecule with the strong mechanical properties and low 
shrinkage of epoxies. The molecule also maintained the fast reaction capability of 
methacrylates [14]. Much of the research since then has evolved around modifications to 
this basic structure. Polymerization in these systems was initiated by a benzoyl 
peroxide/tertiary amine combination. Unfortunately, this initiator system's effectiveness 
was inhibited by the presence of water. Light-cured systems solved this problem as well 
as allowing dentists to work with materials that have not already begun to cure [12]. In 
addition, the chance of incorporating air into the composite system by mixing is 
eliminated. This reduces the number of voids in the restoration [15]. In general, light- 



cured systems cure faster than chemically cured systems [16]. The physical and 
mechanical properties of light-cured systems are better than self-cured [16]. The control 
of setting time and color matching are easier compared to self-cured composites [17]. 
The ultraviolet light systems used initially were discarded for visible-light systems. 
Visible light has better depth of cure than UV as well as the ability to penetrate a layer of 
enamel, which UV light does not [16]. 

Components of Dental Composites 

Dental composites consist of three essential components: a cross-linked polymeric 
matrix, a high volume fraction of particulate silicate reinforcing phase, and a bonding 
agent to promote matrix-filler adhesion [5]. 

Matrix materials are usually based on a combination of dimethacrylates, one 
acting as the base material such as a urethane dimethacrylate, Bis-GMA, or its 
ethoxylated version and another acting as reactive diluent to increase the extent cure as 
well as lower viscosity so higher glass loadings can be achieved. The viscosity of matrix 
portion of commercial composites is typically 1000 to 2000 centipoise (cP) [18]. Taylor 
et al. have evaluated the role of the particle reinforcement volume fraction for an 
extensive range of Bis-GMA-type polymer matrices [18]. It was determined that lower 
viscosity matrices can accept more filler while maintaining workability. Maximum filler 
loading is dependent on the properties of the filler to a larger extent than matrix viscosity. 
An average 4.6 weight percent additional reinforcement could be used in 1000 cp 
solutions compared with 2000 cp solutions. The plasticity of the different matrices 
formulated to the same viscosity and then loaded with the same amount of filler varied 
depending on the chemical structure. The differences were more pronounced at lower 






filler loadings were plasticity varied as much as lOOpercent. The authors refrain from 
hypothesizing about this phenomenon; one could speculate that the interaction between 
different chemical functionalities of the matrix materials and the filler may be one 
possible factor. Reactive diluents such as triethylene glycol dimethacrylate (TEGDMA) 
are effective at lower levels, but large amounts negatively affect matrix properties by 
increasing polymerization shrinkage and water absorption [19]. The polymerization 
shrinkage of poly (TEGDMA) is 1 1.9 percent and it water uptake is 6.0 weight percent. 
The polymerization shrinkage and water uptake of poly (Bis-GMA) are 6.4 percent and 
3.05 weight percent, respectively[5]. 

Filler particles act as the reinforcing phase of the composite as well as reducing 
the overall shrinkage of the system upon polymerization. As long as bonding between 
the filler and the matrix and the quantity of the matrix is sufficient to fill the spaces 
between the filler particles, increased filler content tends to improve mechanical 
properties [5]. Filler particles are usually based on silicates such as zirconium glass, 
yiterbium trifluouride, quartz and bariumborosilicate. The filler materials have been 
developed for stability in the oral environment, aesthetics, and radiopacity. The average 
diameter of the filler particles in traditional composites is about 8 to 30jam, while modern 
composites' average particle size is in the range of 0.7 to 3.6um [20]. Colloidal silica as 
small as 0.04p.m has also been used as filler. Due to the silica's large surface area, i.e., 
300 to 600 m /g, small filler particles increase the viscosity of the composite system they 
are added to, which limits the total volume fraction that can be added [21]. Willems et al. 
showed that the amount of filler in commercial materials ranged from 48.2 to 58. 1 
volume percent [22]. 



8 

The mechanical strength of the composite is significantly enhanced by the use of 
a coupling agent applied to the filler. A coupling agent must chemically bond to the 
reinforcement phase and it must promote wetting by the matrix. However, the long term 
stability of the composite is strongly dependent upon hydrolytic stability of the interfacial 
bonds between the reinforcement phase, coupling agent, and matrix polymer [14]. A 
coupling agent should also be capable of forming covalent bonds with both the matrix 
and the filler. A typical coupling agent is v-methacryloyloxypropyltrimethoxysilane 
(MPS). The silane end is capable of bonding with the filler particle and the methacrylate 
end is capable of forming bonds with the network polymer. 

Areas for Improvement 
Current composite filling materials have also been used for the restoration of 
posterior teeth, but they have not been as successful in this application as for the anterior 
teeth [5]. The main areas for improvement in composite restorations are the following: 
( 1 ) polymerization shrinkage, which results in poor marginal adaptation, internal stresses 
and stress on the tooth; (2) water sorption, which has undesirable effects on dimensional 
and mechanical properties as well as on fillers and filler matrix bonding; (3) thermal 
conductivity and mismatch of CTE with tooth; (4) mechanical properties such as flexure 
and tensile strength as well as modulus; (5) abrasion resistance; (6) adhesion of matrix to 
filler particle as well as matrix to tooth structure and filler to teeth; and (7) radiopacity of 
filler material [5, 21, 23]. 

Network Structure 
It seems that dimethacrylate networks are not uniformly cross-linked. Instead they 
resemble the porridge microstructure first attributed by Houwink to Bakelite [5]. More 



scientifically stated, dimethacrylates are a heterogeneous structure that is characterized 
by areas of highly cross-linked polymer, which surround or contain areas of low cross- 
link density. This is a result of the evolution of structure during cure. Kinetic gelation 
and 2-D percolation modeling indicates that this heterogeneity has its roots early in the 
polymerization process and manifests in the form of isolated branched molecules with 
large numbers of pendant double bonds [9, 10]. Polymerization continues preferentially 
within and around these molecules. These molecules ultimately grow together to form 
the network structure. As a polymer network forms, the onset of vitrification limits 
monomer diffusion, the mobility of pendant groups, and the degree of conversion. The 
vitrification results in areas of low cross-link density that act as defects. These structures 
along with the limited amount of active chains in the network restrict the mechanical and 
physical properties of these materials. 

The heterogeneity of dimethacrylate polymers has been demonstrated 
experimentally using such techniques as dielectric spectroscopy [24, 25], dynamic 
mechanical spectroscopy [26-28], extraction and swelling experiments [29-31], 
photochromic [32], charge-recombination luminescence [33], paramagnetic probes [34], 
CP/MAS NMR 13 C spectroscopy [9, 35] and electron microscopy [36]. There is still a 
need for further characterization of the heterogeneity in dimethacrylates, especially the 
size and distribution of the low and high cross-link density regions. 

Chiu and Lee have isolated prepolymer molecules in ethylene glycol 
dimethacrylate systems [11]. The molecules isolated early in the polymerization had a 
bimodal molar mass distribution (0.4% conversion), i.e., 12 to 18 nm and 35 to 51 nm, 
which was measured by dynamic small angle light scattering (SALS). The number of 



10 

small particles was 3.55 greater than the larger particles, whose structure is comprised of 
clusters of the small particles. Further along in the polymerization the ratio of small 
particles to the larger particles decreased to 2.12, while the average particle size of both 
only increased slightly. At gelation (14.2% conversion), the particle size increased 
considerably. The sizes ranged from 17 to 25 nm and 71 to 173 nm for the small and 
large particles, respectively. The isolated particles had vinyl bond conversions ranging 
from 60-70 percent as by FTIR spectroscopy indicating that microgel particles are 
formed. 

Lange et al. [33] have extended the work on micro-gelled particles using charge- 
recombination luminescence to identify vitrification. Samples cured above the polymer's 
Tg dropped to a baseline value of 2,000 counts during cure and never increased. Samples 
cured at temperatures below the polymer's Tg showed charge-recombination intensities 
(20,000 counts) indicating vitrification from the onset of cure and never dropped to the 
baseline values seen in the sample cured above its Tg. An epoxy recombination 
luminescence spectrum was provided for the comparison of condensation polymers with 
chain growth polymers. The initial portion of an epoxy spectrum mirrored the behavior 
of the dimethacrylate sample cured above its Tg. The intensities increased as vitrification 
point of epoxy was reached. This study provides evidence for the argument that 
vitrification is micro-scale as well as a macro-scale phenomena. 

One of the problems in examining visible light-cured (VLC) systems is that 
radicals are trapped in the microstructure. Techniques that add energy to these systems 
and enhance mobility allow further reactions that can alter the network structure and 
skew results. Kannurapatti et al. [27] have used "living radical" polymerization based on 



11 

iniferters to avoid this problem. The iniferter polymerization proceeds while the system 
is exposed to light. When the light is shut off the reaction ceases, and the structure is 
frozen in either by vitrification or gelation. Dielectric measurements made on poly (Bis- 
GMA/TEGDMA) and poly (EGDMA) [25, 37] reacted with the iniferter system indicate 
an environment with a larger distribution of mobilities. High frequency relaxations 
present are attributed to pools of monomers. Similar relaxations were reported in earlier 
work with triethylene dimethacrylate/methyl acrylate polymers [24]. 

Similar work concerning the nature of networks formed by dimethacrylates 
performed by Simon, et al. [9] elucidated the structures of dimethacrylate polymers by 
cross-polarized proton-enhanced magic angle spinning 13 carbon nuclear magnetic 
resonance (CPPEMAS 13 C NMR) in conjunction with normal 13 C NMR. The CPPEMAS 
technique observes unsaturated groups where isotropic motion is constrained, i.e., 
unreacted chain ends attached to the network structure. The 13 C NMR technique 
observes unconstrained motions, i.e., pools of unreacted monomer. Various results were 
obtained depending on the length and flexibility of the ethylene glycol chain. Systems 
based on tertrakis(ethylene glycol) dimethacrylate reacted to form a network composed 
of only 2 percent pendant double bonds while those based on ethylene glycol 
dimethacrylate were composed of 18 percent pendant double bonds and unconstrained 
monomer. The EGDMA systems showed evidence of pools of unreacted monomer. 
Kinetic gelation models were also used to help interpret the results. In simulations of 
ethylene glycol dimethacrylate polymerizations, gelation models predicted the formation 
of small highly branched molecules early in the cure cycle. The model also predicted the 
growth and eventual coalescence of these molecules resulting in regions of trapped 






12 

monomer. The number of particles was shown to increase from the initiation until 
approximately 5 percent conversion. After 5 percent conversion, the number of particles 
decreases corresponding with an increase in the average size of the particles. This 
indicates that coalescence of these particles is the dominant mechanism of polymerization 
after 5 percent conversion. 

The presence of "Globular Formations" was identified using transmission electron 
microscopy in phenol-formaldehyde resins [36]. The globular formations appeared as 
light and dark regions in the micrograph. The size of dark phases, regions of higher 
density, was determined to be 400 to 600 A. Epoxy, silicone, diallyl phthalate, and 
phenol-formaldehyde samples were also examined by creating replicates of their fracture 
surfaces. The surfaces were examined with scanning electron microscopy. All of the 
samples had globular formations that were on the order of 400 to 900 A. The globular 
structures were rare in the silcone resins when compared to the epoxy, diallyl phthalate, 
and phenol-formaldehyde resins. The identification of the globular phase from the 
replicates of the fracture surfaces should be taken with caution. It is possible the globular 
phases, which appear as bumps in the micrographs, may be artifacts of the fracture 
process. 

Bis-GMA Analogs and Modifications 
Bis-GMA has frequently been modified using the hydroxyl group located on the 
side chain of the molecule. Holter et al. have undertaken a systematic approach to 
modifying Bis-GMA by esterification of the hydroxyl groups with varying molar mass 
aliphatic and aromatic acids [1]. Viscosity and volumetric shrinkage were shown to 
decrease with increasing aliphatic chain length. Unfortunately, modulus was also shown 



13 

to decrease. The aromatic acids were compared with and without an oxy-spacer group. 
The mechanical properties were enhanced by the presence of the oxy-spacer, but the 
viscosity increased. Other studies have shown similar trends in viscosity and cure 
shrinkage by modifying the hydroxyl group of Bis-GMA with dimethylsiloxy and 
dimethylisopropylsiloxy side groups [38]. No mechanical properties were reported for 
these systems. It should be noted when hydroxyl groups are replaced, the system 
viscosities drop one to two orders of magnitude. 

Several works have been published in which the interior propane group of Bis- 
GMA has been replaced by various chemical species [39, 40]. The central propane group 
was replaced with a group containing fluorine and phosphorus atoms with benzene rings. 
Viscosity increased with the stiffness of the central groups. Water uptake decreased in 
fluorine-containing polymers and increased in phosphorus containing polymers. 

Stansbury et al. have extended the work in semi fluorinated monomers by 
synthesizing a wide variety of fluorinated monomers based on Bis-GMA analogs as well 
as a series of urethane-linked multifunctional fluorinated oligomers [40]. The presence 
of fluorine reduced water uptake. The oligomer-based systems have a dry breaking 
strength equal to the Bis-GMA baseline and a comparatively low water uptake. It is 
interesting to note that the highly fluorinated monomers and oligomers had poor wetting 
characteristics of silanated glass filler. The phenomenon is attributed to the relatively 
low hydrophilicity of the matrix materials. The issues of possible reaction kinetics 
differences were not addressed. 

Shoba et al. have synthesized a series of novel dimethacrylates based on 
diphenols with one or two cyclic groups as the central structure (Figure 2.2) [4]. The 



14 

monomers were designed with a variation of straight and flexed structures. All of the 
monomers have the same side chains, which contained methacrylate groups with propoxy 
spacers. The flexed molecules have lower viscosities than the straight molecules. The 
bicyclic monomers have higher viscosities than single-ringed structures. This is due to 
the bicyclic rings making the molecule a stiffer straighter structure. 

In separate papers Sandner et al. and Kawaguchi et al. have reported on high 
molar mass analogs of Bis-GMA [38, 41]. Both papers reported a reduction in shrinkage 
for the higher molar mass polymers. Mechanical properties were variable depending on 
the number of hydroxyl groups in the structure and segmental mobility. Table 2.2 
contains the mechanical properties of various aromatic dimethacrylate polymers. 

Chowdhury et al. showed that binary mixtures of urethane-based monomers with 
three and four methacryl groups have improved compressive and diametral tensile 
properties and increased modulus as measured by micro-indentation compared to a 60/40 
weight percent Bis-GMA/TEGDMA system. Lower amounts of residual monomer were 
also reported [42]. 

Mitra has also explored multifunctional methacrylate monomers [43]. 
Triisocyanato-isocyanurates were used as starting materials in a synthesis that resulted in 
trifunctional ethylenically unsaturated carbamoyl isocyanurates. Increases in both 
compressive and diametral tensile properties were reported for composites based on these 
molecules. 

Oligomers have been created based on poly (isopropylidenediphenol) (BPA)[44]. 
Ethoxylation and methacrylation of the BPA oligomer form these so-called 
multimethacrylates. The multimethacrylates are compatible with TEGDMA. A system 



15 



O 



nr~@-£^ 




13 OH 

Figure 2. 1 : The structure of 2,2-bis(4-(2-hydroxy-3-methacryloyloxyprop- 1 ■ 
oxy)phenol)propane (Bis-GMA) 




A t 




R -\ M\ //- R 



\ / 



R 



R R 



\ /r\ / 



R 



\ / 



CH 3 CH 3 

II 
R = 0— CH 2 COCC=CH 2 

H O 



Figure 2.2: Dimethacrylates synthesized from various diphenols 









16 

composed purely of multimethacrylates had a polymerization shrinkage of 9.2 percent 
compared to 10.6 for a Bis-GMA system. The compressive strengths are higher (477.6 
MPa ) in the multi-methacrylate systems than the Bis-GMA/ TEGDMA systems (375.6 
MPa ) while the diametral tensile strengths are lower (32.8 MPa) compared to Bis- 
GMA/TEGDMA systems (26.4 MPa). No data was presented on the viscosity of these 
monomers. 

Monomethacrylates 
Some of the earliest work with diluents examined a series of mono, di, and tri functional 
methacrylate monomers with a range of molar masses. The diluent monomers were 
incorporated into Bis-GMA-based composites cured with benzoyl peroxide/N, N- 
dimethyl-p-toluidine [45]. Similar mechanical properties were obtained with mono, di, 
and trifunctional monomers. Mechanical properties were more affected by the molar 
mass of the monomers than their functionality. Tensile and compressive strength 
decreased with the increasing length of the linear pendant group on the methacrylates. 
Although methyl methacrylate/Bis-GMA systems showed excellent mechanical 
properties, methyl methacrylate 's low molar mass caused excess shrinkage upon 
polymerization. Estimates or measurement of the glass transition in these polymers 
might provide more insight into the results seen in this work. 

Later work with heterocyclic methacrylates demonstrated that high molar mass 
monofuctional methacrylates could be incorporated with positive effects on mechanical 
properties [46-48]. Tetrahydrofurfuryl methacrylate (THFMA)/Bis-GMA mixtures in 
the volume percent ratio of 5/95, heat cured at 80°C, were reported to have modulus as 
high as 4.8 GPa derived from flexure testing [46]. Bis-GMA tested in the same manner 



17 

was reported to have a modulus of 3.6 GPa. The authors attribute the increase to the 
smaller diluent molecules occupying free volume spaces within the main polymer 
network, thus enhancing the molecular interactions in the polymer. The addition of 
larger amounts of THFMA resulted in lower modulus and flexure strength values varying 
from approximately 80 to 100 MPa. Unfortunately, the statistical significance of these 
mechanical properties tests was not presented. Other work on the same systems showed 
that THFMA did not have significant effects (Newman-Keuls test at a = 0.01) on the 
mechanical properties of Bis-GMA-based polymers when added in portions up to 30 
weight percent [47]. The difference is attributed to varying properties of the Bis-GMA 
monomer due to different isomer compositions. The standard deviation of the strength 
values was approximately 20 percent and none of the values from the different 
compositions were significantly different at an a=0.01. These results suggest the need 
for better testing methods to evaluate the mechanical properties of these materials. 
Fracture toughness measurements would also provide more insight about the strength of 
these materials with less noise due to sample preparation. The monomers 
tetrahydrofurfuryl methacrylate (THFMA), hydroxypropyl methacrylate (HPMA), and 
isobornyl methacrylate (IBMA) were compared to ethylene glycol dimethacrylate 
(EGDMA) and triethylene glycol dimethacrylate (TEGDMA) as alternate diluents in heat 
cured (70°C) Bis-GMA-based polymers. All the monomers reduced polymerization 
shrinkage and increased mechanical properties in the dry state. Equivalent properties 
were maintained in the wet state when compared to Bis-MEPP/TEGDMA polymers. 
HPMA and THFMA were effective viscosity modifiers while IBMA was not. Flexure 
strength was not reported in this study. It is interesting to note that homopolymers of 



18 

IBMA are extremely brittle (10 MPa average flexure strength and 0.71 percent strain to 
failure) compared to those of THFMA, which had an average flexure strength of 94 MPa 
and a 6.88 percent strain to failure [49]. Flexure strength values and glass transition 
temperatures of the cured systems would provide insight into the effects of these 
monomers on network structure and formation. The mechanical properties and glass 
transition temperatures of the homopolymers formed from the diluent monomers are also 
valuable information when interpreting the effects of these monomers. The 
understanding of the effects of these molecules on Bis-GMA type polymers is in its 
infancy and further analysis is needed. The incorporation of alternate mefhacrylate 
monomers with higher molar mass and variable glass transition temperatures into light- 
cured systems would add to the understanding of these effects. 

Polymerization Shrinkage 
Methacrylates exhibit a volume reduction, from 12 to 22 percent [50, 51] or 22.5 
ml/mol [52], during polymerization which is associated with two phenomena: (1) the 
change in intermolecular distances of 3 to 4 angstroms to primary covalent bonds lengths 
that are 1.5 angstroms [6]; and (2) the increased packing density of polymer chains 
compared to the packing density of monomer molecules [1]. Many restoration failures 
are associated with this volume contraction. The problems that result are stresses that are 
large enough to cause defects within the matrix and debonding at the restoration/tooth 
interface, which undermines the performance of the restoration [7]. The problem of 
shrinkage has been addressed in many ways. The addition of inorganic reinforcement 
phase is commonly used to lower the overall shrinkage in composites by reducing the 
volume fraction organic materials present [23]. Many different sizes of filler as well as 



19 

mixtures of several different sizes, referred to as hybrids, have been used to maximize the 
amount of filler incorporation. 

Liu et al. have developed an ammonia-modified/hydrated mineral 
montmorillonite (NH3/MMT) filler [6]. The exotherm of cure vaporizes the ammonia, 
which swells the filler and offsets polymerization shrinkage. This system results in zero 
shrinkage with small additions of filler (4 weight percent). However, NH3/MMT 
requires large volumes of polymerizing material to create a sufficient exotherm to drive 
the expansion reaction (60-80°C), which is more than a sufficient amount of heat to 
damage tissue. The mechanical properties were not affected drastically due to the small 
volume fraction of the filler. An SEM examination in the same study revealed no 
porosity. However, the electron beam degraded the sample rapidly, which makes these 
results suspect. It was claimed that reduction in volumetric shrinkage was due to the gas 
that evolved during the reaction and was trapped in the filler particles. There is no study 
available that addresses the issue of long-term gas release and its effect on the porosity of 
the polymer. However, one might suspect that the diffusion of the gas from the network 
could enhance diffusion into the pores by fluids. This presence of the fluid would reduce 
the integrity of the composite. 

Another common approach has been to increase the specific molar mass of the 
monomers used in the system. Several groups have used this approach with urethane 
dimethacrylates [41, 42]. The same approach has been used in the synthesis of higher 
molar mass analogs of Bis-GMA [38, 41]. Culbertson et al. has formulated higher molar 
mass multi methacrylates based on an oligomer chain of bis-phenol A molecules [44]. 
Bowen has also developed a molecule based on beta-cyclodextrin [53]. Shobha et al. 



20 

created Bis-GMA analogs with bent structures as well as varying degrees of stiffness [4]. 
The stiffer, straighter structures had the least shrinkage. 

In an attempt to control the density of chain packing, several studies have 
modified the structure of Bis-GMA through the hydroxyl groups [1, 54]. Kalachandra 
replaced the hydroxyl groups with trimethyl and isopropylsiloxyl groups [54]. The bulky 
isopropyl group reduced shrinkage to a greater extent. No mechanical testing was done. 
In a more systematic and in-depth study, Holter esterified the hydroxyl groups with a 
series of aliphatic and aromatic acids [1]. Polymerization shrinkage and modulus trends 
are shown in Table 2. 1 . The aromatic acid-modified Bis-GMA molecule had comparable 
reductions in shrinkage, but less reduction in modulus. This was especially true where a 
flexible oxy-spacer group is used. This work indicates that shrinkage may be reduced 
without drastic losses in mechanical properties using structure-property relationships. 

Liquid crystal molecules have also been studied in an attempt to minimize 
polymerization shrinkage [55]. The ordered nematic structure is denser than a typical 
monomer solution and the order-to-disorder transition that occurs during curing results in 
a less dense system. The nematic state also results in lower viscosities; therefore, higher 
molar mass monomers can be used further increasing the system's potential for offsetting 
shrinkage. The use of acrylates brings up concerns of toxicity, but future work is planned 
with methacrylates. 

High molar mass analogs of Bis-GMA monomers can only reduce polymerization 
shrinkage. A monomer that expands upon polymerization is necessary to have a system 
that expands or has no volume change on polymerization. Stansbury and Byerley et al. 
were the first to attempt to develop spiro orthocarbonates (SOC) for incorporation into 



21 

dental monomers to offset polymerization shrinkage [7, 56]. Spiro orthocarbonates offset 
shrinkage by a double ring-opening free-radical polymerization mechanism. The first 
SOCs developed were crystalline and not compatible with monomer systems. Stansbury 
and Byerley et al. have attempted to rectify these problems in their early work. Byerley 
et al. has developed a series of stereoisomeric alicyclic spiro orthocarbonates, which 
expand between 3.9 and 3.5 percent upon polymerization. Low molar mass polymers 
were produced by a light-initiated cationic polymerization of the experimental 
monomers. All of these monomers, none of which were tested in dental monomers, are 
crystalline at room temperature, which would most likely decrease their solubility in 
dental monomer systems. Stansbury has taken the concept of SOCs and synthesized 
monomers with unsaturated functional groups that should allow the incorporation of the 
monomers with free radical-type systems. The systems were cured using photo, 
chemical, and dual cure mechanisms. Diametral tensile strength (DTS) values of 



Table 2. 1 : Volumetric shrinkage values and modulus values for Bis-GMA molecules 
esterified with various- length aliphatic acids. The number next to C represents the 

number of carbon atoms in the aliphatic chain 

Compound Volumetric Modulus 

shrinkage (MPa) 

(%) 



Source: Holter et al. [1] 



Bis-GMA 


4.3 


2100 


Bis-GMA-C2 


4.7 


1500 


Bis-GMA-C6 


5.0 


220 


Bis-GMA-Cll 


3.9 


- 


Bis-GMA-C18 


2.6 


- 



22 

44.4 ±2.1,45.1 ± 4.6, and 50.4 ± 2.8 (MPa) were reported for 100 percent Bis-MEPP, 
65/35 weight percent mixture of Bis-MEPP and 1,6 hexanediol dimethacrylate, and 68/32 
weight percent mixture of Bis-MEPP and SOC monomers. This result may be 
misleading because the SOC molecules may be acting as a plasticizer. Some 
compositions were not stable and underwent polymerization during storage while others 
did not polymerize fully [7, 56]. There was also a lower offset of polymerization 
shrinkage than expected, which is attributed to an alternate ring opening polymerization 
system. Reaction kinetics and volume control under non ideal curing conditions are also 
limiting factors in these systems. 

SOC molecules were incorporated into epoxy resins in later work based on the 
initial work of Byerley et al. [57, 58]. The use of polyols was explored to catalyze the 

* 

polymerization. The systems were initiated with onium salt initiators and 2- 
chlorothioxanthen-9-one (CTX). The addition of SOCs to the composition caused an 
increase in cure times over the baseline epoxy resins. The fastest cure time reported for a 
mixture that contained a SOC was 48 minutes. The introduction of polyols reduced the 
cure times; however, no studies were presented with both polyols and SOCs present. It is 
possible that the polyols will open the SOC prematurely. 

The advent of low toxicity onium light cure initiators has generated renewed 
interest in epoxy resins for dental applications. Tilbrook et al. have addressed the 
problems of water uptake in epoxy-polyol mixtures by varying the ratio of epoxy groups 
to polyol groups. The samples were cured using a camphorquinone/4-octyloxy-phenyl- 
phenyl iodonium hexafluoroantimonate system. The camphorquinone results in a system 
that is sensitive in the blue light regime commonly used in dental composites. The 



23 

polymerization shrinkage for epoxy systems formulated ranged from 3.8 to 5.4 percent. 
The water uptake ranged from 5.2 to 7.7 weight percent. Unfortunately the resins with 
the lowest amounts of shrinkage also had the greatest amount of water uptake. Baseline 
Bis-GMA/TEGDMA systems had less water uptake and higher shrinkage upon 
polymerization. The epoxy formulations showed higher biaxial flexure strengths than 
Bis-GMA/TEGDMA systems and modulus values were also higher for some 
formulations. All samples were tested in the dry state. The uptake of water will affect 
the properties of materials. Comparisons of these materials in the wet state should be 
made before further recommendation is made for implementation of the materials in 
composite restorations. The heat rise in the epoxy samples during cure might also be an 
issue. It was also noted that some samples had curing times in excess of 60 seconds. 
This time would not be acceptable to clinicians. An interpenetrating network of 
methacrylates and epoxies might take advantage of the favorable properties of both 
systems. 

Mosner et al. have synthesized three series of radical ring opening monomers 
[59]. The synthesized molecules include SOC molecules with substituted unsaturated 
groups, as well as bicyclic 2-methylene-l,3,-dioxepane (BMDOE) derivatives, and 1,1- 
bis-substituted 2 vinylcyclopropanes. IR data showed that the SOC molecules 
polymerized through the unsaturated group. The ring opening mechanism was not 
utilized and no offset of shrinkage was seen. The BMDOE molecules synthesized were a 
mixture of liquid and crystalline monomers. Spectroscopic data indicated that ring 
opening takes place during polymerization. The crystalline molecules expanded upon 
polymerization. Some of the volume gain was attributed to the transition from the denser 



24 

crystalline state to the amorphous state during polymerization. Dental applications of 
these molecules are limited by their sensitivity to moisture and acid compounds. The 
BMDOE molecules react to form aliphatic polyesters that have low glass transition 
temperatures (27 to 37°C). The vinylcyclopropanes synthesized were able to form cross- 
linked networks when polymerized in bulk. Depending on the initial morphology, 
volume change during polymerization varied. The molecules that formed amorphous 
liquids shrank 3.9 and 7.0 percent, while the crystalline molecules expanded 1.0 percent. 

Miyazaki et al. have synthesized methacrylates and acrylates that have SOC 
groups [60]. Heat curing resulted in a minimum shrinkage of 7.8 percent compared to 9.8 
percent for Bis-GMA. Studies still need to be performed with a more suitable initiator 
system. The heat cure was performed at 120°C, which delays the onset of vitrification 
and allows higher amounts of conversion. Light initiation systems are used at room 
temperature. At these lower temperatures vitrification will limit the mobility of 
molecules and hence the conversion that can be obtained. The mechanical properties of a 
light-cured system will be significantly different than those of a heat-cured system. 

Bowen has synthesized a methacrylated [3-cylcoldextrin [53]. It is hoped that by 
housing monomer in the hydrophobic cavities of these molecules, polymerization 
shrinkage will be offset. If the monomer contained in the cavity diffuses out during 
polymerization and becomes an external chain segment, then the additional free volume 
created might help offset shrinkage. The issues of how the molecules will be placed into 
the cavity and controlling their diffusion properties have yet to be addressed. 

Many approaches have been taken to minimize or offset the shrinkage that occurs 
upon polymerization. As yet, a non-shrinking system has not been introduced that meets 



25 

all the requirements for use in the mouth. Reduction of polymerization shrinkage has 
been obtained with molecules that have a larger molar mass than Bis-GMA. Ultimately 
these systems will offset polymerization shrinkage modestly and viscosity will become a 
limiting factor for large molar mass molecules. Liquid crystalline molecules show 
promise, but the toxicity of the current molecules and their ability to form a LC phase at 
lower temperatures are still issues. Ring open molecules must be employed to have zero 
shrinkage or an expanding system. Unfortunately, these molecules are hampered by slow 
reaction kinetics, moisture sensitivity, and crystallinity. 



Anhydrides in Dental Materials 

Numerous studies have evaluated the use of anhydrides as comonomers for both 
structural and adhesive purposes in dental materials [8, 61-66]; however, none have 
attempted to capture the ring opening as a mechanism for offsetting or reducing 
polymerization shrinkage. The anhydride functionality may also allow tailoring of 
mechanical properties through ionic interaction as well as a viable means to increase 
bonding to bone tissue. 

A series of anhydrides were added as cross-linking agents to polymer composites 
based on UDMA, HEMA Bis-GMA, and TEGDMA [8]. It was hypothesized that 
monomers such as maleic anhydride introduced into dental materials containing hydroxy 
or amide groups would act as cross-linking agents. A cyclic anhydride can react with 
either of these functionalities and form an ester or amide linkage upon additional heating. 
This leaves the hydroxy group at the end of the carboxylic acid to react with a second 
amide or hydroxy functional group and form a di-ester or di-amide linkage. The best 



26 

results were found with anhydrides that contained unsaturated groups. It was suggested 
that the double bond makes it possible for these anhydrides to copolymerize with the 
other monomers. This seems counterintuitive, as polymerization with the network would 
lower the mobility of these species and the average functionality of these systems. The 
anhydride incorporation might also lower the cross-link density of these systems. In the 
case of maleic anhydride, for example, polymerization of the double bond may reduce the 
ring strain and hence the reactivity of the anhydride group [67]. Polymers based on 
UDMA and HEMA had more desirable properties. This is attributed to the greater 
reactivity of the amide groups with the anhydrides, hence a greater degree of cross- 
linking. A 20 percent increase in mechanical properties was reported for the combination 
of maleic anhydride and methacrylamide. There appears to be a synergistic effect 
between these two monomers. The authors provide no explanation for the observed 
effect. The results obtained in this work are interesting and suggest that the use of 
anhydrides as cross-linking agents may have some validity. Later studies done on 
selected anhydride containing systems from the initial work examined the relationship 
between wear and the quantity of remaining double bonds [63]. The effect of post cure 
heating on mechanical properties was also evaluated [62]. In vitro wear decreased with 
increased conversion of double bonds with the exception of the samples post cured at 
37°C and 75°C. There was no difference in the quantity of remaining double bonds 
between these two samples; however, the wear rate was lower in the sample cured at 
75°C. It was suggested that this is the result of cross-links formed by anhydride 
molecules. Clearly no spectroscopic evidence is provided about whether or not this is the 
mechanism responsible for the reduction in wear. 



27 

The ability of anhydrides to be incorporated into dental polymers was 
demonstrated with the model compound 2-phenylethyl methacrylate and propoxylated 
Bis-GMA/TEGDMA [68]. The ability of anhydrides to offset polymerization shrinkage 
was confirmed by post polymerization volume expansions observed after swelling in 
water as high as 2.0 percent. The hydrolysis was confirmed by a residual weight gain in 
dehydrated samples and FTIR observations. 

One of the most important dental applications for functionalized monomers is 
their use as dentin adhesives [65]. Anhydrides are of interest in dentin bonding systems 
and bone cements because of their reported complexation with calcium cations. The 
addition of 4-methacryloyloxyethyl trimellitate anhydride (4-META) to standard MMA 
bone cements has been shown to increase adhesion to bone and hydroxyapatite [65, 66]. 
Increased bond strengths to dentin and enamel have been shown when 4-META and 
pyromellitic dimethacrylate (PMDM) were applied as coupling agents prior to bonding 
with composite polymers [61, 69]. Bond strength from 10.2 to 18.2 MPa were reported 
when PMDM and 4-META were used in conjunction with ferric oxalate solutions and 
N(p-tolyl)glycine glycidyl methacrylate (NTG-GMA) or N-phenylglycine glycidyl 
methacrylate (NPG-GMA) surface active monomers. The bond strengths reported for the 
composites used with the ferric oxalate ranged from 0.48 to 1 .72 MPa. Shear bond 
strengths of composite to human dentine of 29.7 ± 1 1.8 MPa have been reported when 
the sodium salt of NTG-GMA was used as a bonding agent[70]. Maleic anhydride 
dissolved into standard dental monomers has also been studied as an adhesion promoter 
for Cr-Co alloys [64]. Unfortunately, initial improvements in the energy of adherence 



28 

were lost after storage in water for two months, most likely as a consequence of water 
adsorption. 

Water Absorption 

Water in dental materials induces swelling, leaches extractable material out of 
both the polymer and the filler, plasticizes polymers causing them to lose their 
mechanical properties, and degrades the bond between fillers and coupling agents. There 
are many variables that govern the uptake of water, and no one has been able to sort them 
all out yet. Hydrophilicity is related to the uptake of water. The hydrophilicity of a 
material seems to be based on how similar the polymer structure is to water. Polar 
materials are more hydrophilic. Kalachandra et al. have found a linear correlation 
between water uptake and weight percent oxygen (WPO)[71]. WPO is a more concrete 
way of defining hydrophilicity in polymers. Interestingly, the trend breaks down when 
there is a branch in the monomer that sterically hinders the polar sites from the water. 
Bis-GMA-type monomers also showed less water uptake than expected based on WPO. 
It was hypothesized that the reduction in water uptake is due to the formation of an 
intramolecular hydrogen bond between the carbonyl oxygen and the hydrogen of the 
secondary hydroxyl group. The intramolecular bond reduces water uptake by blocking 
the hydrophilic groups and acting as an additional cross-link. 

The amount of free volume in the sample also has an effect on the amount of 
water taken up and hence the amount of swelling. Soderholm et al. have shown that 
polymers do not swell in a 1 to 1 ratio with the volume of water absorbed [72]. The lag 
in the swelling of the polymer was attributed to microcavities that accumulate water. The 
structure of the network formed clearly has an effect on water uptake. It has been shown 



29 

that there is a relationship between water uptake and degree of cross-linking [73]. As the 
dimethacrylate feed was increased into a PMMA polymer, so did the uptake of water. 
Although the cross-link density of the materials increased with dimethacrylate feed the 
number of pendant groups, trapped unreacted material, and regions of lower cross-link 
density also increase. This may account for the increase in water uptake. Simon et al. 
have proposed that solvent uptake occurs preferentially in regions of lower cross-link 
density [30]. In contrast to this theory, the water uptake of ethylene glycol methacrylate 
monomer is about 2percent less than its polymer. It might be possible that interaction 
between the various chemical groups prevents their interaction with water thus, lowering 
water uptake in the monomer systems. In the polymeric state mobility restrictions might 
prevent these interactions, therefore chemical groups might be more accessible to water 
molecules. As discussed earlier such an interaction was attributed to the lower than 
expected water uptake of Bis-GMA polymers. The physical phenomena underlying these 
results are difficult to ascertain. 

Mechanical Properties 
Adequate mechanical properties are necessary for a restoration to function 
properly. For example, a restoration with low modulus will readily deform elastically 
under stress. This deformation could result in failure of the tooth around the composite 
as well as increased leakage [21]. The restoration itself might also fail if it has low 
modulus or fracture strength. An understanding of the mechanisms that govern 
mechanical properties of filled and unfilled systems both wet and dry is important so 
improvements can be made in a systematic fashion rather than randomly. 





30 






Kawaguchi et al. have done extensive work relating the basic structure of 


dimethacrylates to their mechanical properties [41]. 


Nine aromatic dimethacrylates 


(Figures 2.3-2.5) were 


studied and their wet and dry mechanical properties 


were 


discussed. All of the monomers were mixed 50/50 by weight percent with 


triethylene 


glycol dimethacrylate 


(TEGDMA). Table 2.2 summarizes the results from the 


mechanical properties 


tests. Water sorption data is also included. 




Table 2.2: Mechanical properties and water sorption 


of various unfilled dimethacrylate 


polymers wet and dry 








Monomer 


Flexure 


Elastic 


Amount of 


Composition 


Strength* 


Modulus* 


Water 




(MPa) 


(GPa) 


(p.g/mm 3 ) 


Bis-GMA/ 


Dry 149.1+7.3 


3.5310.05 


58.910.03 


TEGDMA 


Wet 99.714.6 


2.1510.06 




Bis-GMA-F/ 


Dry 156.7±6.6 


3.7410.09 


55.410.04 


TEGDMA 


Wet 119.6±6.6 


2.8710.07 




Bis-GMA-C/ 


Dry 101.112.6 


2.5110.05 


62.310.02 


TEGDMA 


Wet 80.312.4 


1.2210.04 




Bis-GMA-Eph/ 


Dry 101.713.2 


2.6410.10 


35.410.10 


TEGDMA 


Wet 61.411.3 


1.5010.05 




Bis-GMA-Ech/ 


Dry 113.413.2 


2.8510.11 


40.210.04 


TEGDMA 


Wet 70.215.1 


1.8310.08 




Bis-MEPP/ 


Dry 170.317.3 


2.9610.06 


20.510.04 


TEGDMA 


Wet 123.816.1 


2.8910.07 




Bis-MEEPP/ 


Dry 96.612.4 


2.1910.06 


22.510.05 


TEGDMA 


Wet 93.811.5 


1.3110.02 




BisMPPP/ 


Dry 113.114.3 


2.4810.04 


16.810.10 


TEGDMA 


Wet 107.613.4 


2.3310.04 




BisME 2 . 6 PP/ 


Dry 120.714.6 


2.7210.05 


18.210.02 


TEGDMA 


Wet 115.712.4 


2.57 10.08 




*Measured by three-point bend testing on a universa 


testing machine 




Source: Kawaguchi et al. [41] 















31 

One of the most obvious trends in this data is the loss of wet mechanical 
properties in the polymers with hydroxyl groups. The polymers without hydroxyl groups 
retain mechanical properties after soaking in water. This phenomenon is associated with 
plasticization by water and is supported by the correlation seen between water uptake and 
decrease in strength. Bis-GMA-EPh and -ECh are high molar mass analogs of Bis-GMA. 
Their mechanical properties are lower than those of Bis-GMA. The lower values of the 
mechanical properties are attributed to higher segmental mobility. It should also be noted 
that these two compounds are more hydrophobic than Bis-GMA, which resulted in less 
strength loss under wet conditions. The more hydrophobic nature of the molecules is due 
to the smaller amount of polar groups per molar mass. This also lowers modulus values. 
The fluorine-containing Bis-GMA polymers had excellent properties in both wet and dry 
states. Fluorine groups make the monomer more hydrophobic, which is seen in the lower 
uptake of water. The Bis-GMA-C monomer absorbed significantly more water than the - 
EPh and -ECh versions, yet maintained a similar amount of strength. More information 
is needed about the network structure of these monomers to hypothesize about these 
results. The non hydroxyl containing monomers all have good retention of strength when 
tested under wet conditions. Again, the mechanical properties can be related to 
segmental mobility, in this case, of the aliphatic side chains. The monomers with longer, 
more flexible side chains have lower mechanical properties. The addition of stiffer 
molecules does not always improve mechanical properties. TEGDMA, a more flexible 
and less viscous molecule than Bis-GMA, is added to all of the compositions discussed in 
this paper. The addition of TEGDMA allows further reaction of the system before the 
onset of vitrification. 






32 

Ferracane has shown that if enough TEGDMA is not present that the consumption 
of double bonds by polymerization or degree of conversion (DC) can be limited (Table 
2.3). The stiffer, more viscous Bis-GMA molecules limit the network formation [3] 

129 Xenon Nuclear Magnetic Resonance Spectroscopy 
Xenon Nuclear Magnetic Resonance spectroscopy (NMR) has become an increasingly 
popular technique for sampling a wide variety of materials because of the sensitivity of 
xenon's chemical shift to its environment. Xenon, referred to as xenon from here 
forward, environmental sensitivity is due to its polarizable electron cloud. Xenon's 
relatively small size and chemical inertness also make it attractive as a probe molecule. 



Table 2.3: Degree of conversion values for Bis-GMA/TEGDMA polymers systems 

o btained by Ferracane 

Monomer (wt.%) Catalyst (wt.%) DC (%) 

75/25 0.25 DMPT + 0.75 BPO 60.1±1.9 

+ 0.01 BHT 
75/25 0.3 1 DMPT + 0.75 BPO 60.5±2.0 

+ 0.01 BHT 
75/25 0.20 camphoroquinone + 60.1+2.7 

0.75 DMAEM + 0.01 

BHT 
50/50 0.25 DMPT + 0.75 BPO 69.4+1 .3 

+ 0.01 BHT 
50/50 0.20 camphoroquinone + 71.5±0.9 

0.75 DMAEM + 0.01 

BHT 



Note: BPO - benzoyl peroxide, DMPT = N, N-dimethyl-p-toluidine, DM APE - 4-N,N,- 
dimethylaminophenethyl alcohol, BHT = butylated hydroxytoluene (inhibitor), DMAEM 
= N,N-dimethylaminoethyl metharcrylate. The coefficient of variance for the modulus 
values is 7 percent. Source: Ferracane [3] 



33 



O 



O 



OH N ' N ' 




or >r o 




R = C(CF3)2 : Bis.GMA F 



:Bis GMAC 



OH > ' CH, OH 



B 






R = 




Bis GMA EPh 



:Bis.GMAECh 






CH 3 
I 
B = 0-CH 2 CH20CC=CH 2 

O 

Figure 2.3: Bis-GMA analogs referred to in Table 2.2 



34 









o 



o 




m 




CH 3 
C 





CH 3 




() 



m = n = 1 : Bis MEPP 

m = n = 2 : Bis MEEPP 

m + n >= 2.6 : Bis.ME 2 .6PP 
Figure 2.4: Ethoxylated Bis-GMA molecules 



,0 



n 




O 




o^° 




Figure 2.5: 2,2'-bis-(4-methacryloylpropoxyphenyl) propane (Bis-MPPP) 



35 

The nature of the chemical shift of xenon molecules has been assessed and the major 
portion of the shift is due to Van der Waals dispersion and repulsive interactions [74]. 

Initially xenon was used to examine the porous structure of solid materials such as 
zeolites, clathrates, and hydrates. The first applications of xenon NMR to polymers 
examined properties of the amorphous portions of bulk polymers. Sefcik was the first to 
report a xenon NMR signal in PVC. A broad signal, which is typical of polymers above 
their glass transition, was observed [75]. Stengle and Williamson later examined poly 
(ethylene) and poly (ethyl methacrylate) and elucidated that the xenon chemical shift is 
sensitive to the density of its solvent system, different amorphous environments with the 
same polymer, and glass transitions [76]. The sensitivity of xenon to the density of its 
solvent system was demonstrated using a series of N-alkanes. The chemical shift of 
xenon varied linearly with increasing molar mass of the alkanes. The shift ranged from 
160 ppm for hexane to 200 ppm for LDPE. The increase in molar mass correlated with 
an increase in density. The densities in the systems ranged from 0.65 to 0.87 g/cm 3 . The 
glass transition of poly (ethyl methacrylate) correlated to changes in line width and the 
chemical shift of the xenon NMR signal. A single peak was observed for LDPE while 
LLDPE showed a partially resolved doublet. The second peak indicates the presence of 
two different sub regions within the amorphous phase. Solubility within the crystalline 
phases of both polymers was assumed to be negligible. This is reasonable when the later 
work of Kentgens et al. showed that the signal to noise ratio decreased as the amount of 
crystallinity increased [77]. 

Kennedy identified 4 phases in EPMD rubber with xenon NMR[78]. The 
different peaks are associated with regions of different void space. Upon cross-linking 



36 

the peak that was furthest upfield, associated with the regions with the most void space, 
disappeared. The magnitude of the other peaks changed dramatically. This work 
illustrates the ability of xenon NMR to monitor changes in structure. 

The line width of the xenon NMR signals is related to the distribution of 
absorption sites within a polymer and the frequency at which the xenon samples or 
exchanges between these sites. The three types of exchange are fast, slow, and 
intermediate [79]. The fast exchange occurs when the exchange rate of the xenon is rapid 
compared to the NMR time scale. Intermediate exchange occurs when the exchange time 
is on the order of the NMR time scale. Slow exchange is observed when the rate of 
exchange is slow compared to the NMR time scale. The NMR time scale is the time in 
which the NMR signal is collected. The length of sampling time is often dictated by the 
T2, spin-spin, relaxation times. Two-dimensional NMR can be used to confirm which 
type of exchange is taking place. Kentgen et al. performed such experiments with poly 
(ethylene) and poly (carbonate). The 2D poly (carbonate) spectra is round and almost 
2000 Hz wide. The round shape indicates that the xenon NMR is sampling the different 
absorption sites in a slow exchange regime. Slow exchange is probably the case for most 
polymers below their glass transition temperature. Polymers such as poly (ethylene) have 
narrow line- widths on the order of 100 Hz. The narrow width is indicative of fast 
exchange. The chemical shifts of the different sites average and a narrow peak results. 
Kentgens et al. [77] tested poly (ethylene) at a range of temperatures and showed a 
dramatic increase in line width as the temperature of the P-relaxation temperature was 
approached. This supports the theory that narrow peaks in polymers above their glass 
transition are due to signal averaging. The mobility of polymer chains above their glass 






37 

transition is great. This allows for rapid xenon movement and thus fast exchange. Below 
their glass transition temperature the motion of polymer chains is curtailed, the diffusion 
of xenon gas is slowed and no averaging takes place, thus a broad peak is observed. 

Xenon NMR has also seen limited use a means to study free volume in polymers. 
Chu et al. [80] have studied the relaxation of free volume in polymers below their glass 
transition temperature. The chemical shift of xenon continued to shift for several weeks 
after the polymer had been cooled below its Tg, indicating the free volume of the 
polymer continued to change. Morgan et al. have studied free volume distributions in 
dendritic and cross-linked polymer systems derived from poly (oxypropylene) diols and 
2,4-toluene diisocyanate [81]. The xenon chemical shift increased linearly with 
increasing cross-link density. In variable temperature experiments signal intensity 
decreased as the cross-linked systems were cooled through their Tg. The decrease in free 
volume associated with the polymer's glass transition lowers the amount of absorbed 
xenon. The chemical shift of xenon was examined in dendritic structures in various 
generations. The chemical shift of xenon decreased linearly when compared to the 
number of chain ends normalized by the average molar mass of a chain. It is suggested 
that there is a relationship between generation number of the dendrimer and free volume. 
It is also noted that free volume measured is an average for each macromolecule and that 
there may be distribution of free volume within the molecule. 

Xenon's most useful application may be as a tool determining miscibility in 
polymer blends and measuring the size of morphological features in polymers and 
polymer blends. It has been estimated that Xenon NMR techniques should be able to 
measure domain sizes from 0.1 to 25 micrometers [79]. Generally, if multiple phases are 



38 

present then multiple peaks will be observed in the NMR spectrum. This is due to the 
sensitivity of xenon to its chemical environment as discussed earlier. In the case of small 
domain sizes, a single peak can be observed due to rapid exchange between the domains. 
This problem can be addressed by using variable temperature experiments because the 
diffusion rate of xenon is temperature dependent [79]. 

Xenon has been used as a probe molecule in many polymer blends [82-89]. 
Browstein et al. were the first to study polymer blends using xenon NMR Two signals 
were observed in poly (styrene) poly (isoprene) blends [89]. The peaks were assigned 
based on linewidths. It was noted that the resonances were broader than in the pure 
materials. The broadening was attributed to the exhange of xenon between the two 
phases. 

Walton et al. demonstrated that xenon NMR can be used as a tool to determine 
phase segregation in blends of poly (chloroprene) PC, poly (isoprene) PIP, and 
epoxidized poly (isoprene)[84]. Blends of PC/PIP, 50 percent epoxidized poly 
(isoprene)/PIP and 25 percent epoxidized poly (isoprene)/PIP, all known to be 
immiscible, showed two distinct peaks in the xenon NMR spectrum. The location of the 
peaks matched those of the pure polymers. A single resonance is observed in PC/50 
percent epoxidized poly (isoprene) and PC/25 percent poly (isoprene) blends. A single 
resonance does not unequivocally indicate a miscible blend. It is possible for xenon to 
exchange rapidly between small domains. This would result in a signal that is an average 
of the signals from the two domains. Knowledge of the diffusion coefficients of xenon in 
these polymers yields an upper boundary on the domain size in which rapid exchange 
could take place. The possibility of miscibility was further examined by modeling the 



39 

signal based on rule of mixtures calculations. It was assumed that a miscible blend has a 
different signal than a signal from a blend that is the result of an average from the two 
separate phases. A signal that does not match these predictions is indicative of miscible 
polymer. The possibility of a chance match in the signal of the miscible blend and an 
average signal was ruled out using variable temperature experiments. This application 
demonstrates the usefulness of xenon NMR for determining phase segregation in 
polymers with similar glass transition temperatures. The miscibility of these blends 
could not be determined using DSC because of the similarity of their glass transition 
temperatures. 

Walton et al. have used xenon NMR to study phase transitions in poly (isoprene 
and poly (butadiene) blends[88]. Phase segregation was detected beyond the limits of 
DSC. Sizes of the phases were estimated using the intermediate exchange assumptions 
and the xenon self-diffusion coefficient in poly (isoprene). 

Mirabella and McFaddin examined a poly (propylene) ethylene-propylene rubber 
blend using xenon NMR[85]. A correlation was shown between peak location and 
domain size as measured by SEM. Calibration in this manner might allow domain size 
measurement in other blends of these polymers. The viable range of measurement with 
this technique is limited by xenon's rate of diffusion in these polymers because slow 
exchange must be maintained. 

Tomaselli et al. have demonstrated the ability of two-dimensional NMR 
techniques to probe heterogeneous blends using fabricated lamellar structures. Two 
model systems were manufactured poly (styrene) or poly (vinyl chloride) and poly (vinyl 
methyl ether). The lamellar size for the PS-PMVE systems was 4 to 7 ^im and 2 to 6 um 



40 

for the PVC-PVME systems. Two-dimensional experiments were performed with 
various mixing times allowing the xenon molecules various amounts of time to move 
through the structures. The extent that xenon sampled the different environments was 
determined by examination of the cross-peaks. The ratio of the diagonal peaks to the 
cross-peaks was used to calculate an exchange rate constant. By assuming a lamellar 
size, a diffusion coefficient for the blend was determined using the exchange rate 
constant. There is a large amount of error due to the lamellar size assumption. It is also 
assumed that the rate of diffusion was similar for both polymers. 

Schantz and Veeman used two-dimensional spectroscopy to examine 
heterogeneity in PEO/PMMA blends [87]. A two-dimensional spectrum demonstrated 
that the xenon sampled all the environments in a mixing time of 1 ms. Furthermore, two- 
dimensional experiments were forgone in favor of one-dimensional spectra in which 
exchange-average models were used to interpret data. The authors chose this approach 
because of the long times required for two-dimensional experiments and because the 
xenon was probing the system in the fast exchange regime. The models were based on 
the spectral parameters from the pure polymers. The exchange rate parameter was then 
adjusted so that the model spectra matched the experimental. Domain sizes were then 
estimated based on the exchange rate and an approximate diffusion coefficient. The 
uncertainty in the fitted values results in values ranging from 30 to 60 nm. This work 
demonstrates the ability of xenon NMR to measure domain sizes on the order of 50 nm. 
The need for accurate diffusion coefficients to determination of domain sizes in polymers 
is elucidated. 



41 

Xenon diffusion coefficients in polymers have been determined in several ways. 
Simpson et al. have measured xenon's self diffusion coefficient in poly (styrene) using 
microspheres of different and well controlled sizes. The rate of diffusion of the gas out 
of the spheres was then related to free induction decay of the signal of the absorbed 
xenon. The difference of length of the free induction in the xenon from the different size 
spheres is plotted. The result allows calculation of diffusion coefficients. This technique 
is dependent upon the ability to fabricate microspheres and that the absorbed xenon has a 
relatively narrow peak. Tomaselli et al. measured the average diffusion coefficients for 
two polymers in a similar way by manufacturing artificial lamellar structure and using 
two-dimension exchange experiments as mentioned earlier. Browstein et al. made also 
calculation of the diffusion coefficient in poly (styrene)/poly (isoprene) blends using 
calculated domain sizes and spectral line-widths [89]. It was also shown that the line 
width of the xenon NMR signal was related to the diffusion of the xenon from one phase 
to another. The effect of the interphase between the blocks was not considered. 

Perhaps a more elegant technique of measuring the self-diffusion of xenon is the 
pulsed field gradient technique. Stejskal and Tanner developed this technique in 1965 
using proton NMR [90]. Junker and Veeman have applied it to xenon in polymers using 
a stimulated echo sequence rather than a Hahn echo [86]. The stimulated echo sequence 
aids in signal collection, which is difficult with xenon absorbed in polymers due to its 
short T2, spin-spin, relaxation times. The diffusion coefficient of the elastomer ethylene- 
propylene-diene was shown to be approximately 20 times greater than that of poly 
(propylene). The pulsed diffusion technique is limited by the long times necessary to 
collect signals. The lengths of experiments in this study were on the order of 70 hours. 






42 

The long collection times are due to the short T2 times (3-8 ms), low signal to noise 
ratios, and long Tl times. 

Dynamic Mechanical Spectroscopy 

Dynamic mechanical spectroscopy (DMS) is a widely used technique for 
characterizing polymers but it has found limited applications in dental materials. Dental 
restorations experience a wide variety of temperatures in their service life. The ability of 
DMS to test properties at a variety of temperatures and provide information about 
network structure is desirable. 

Some of the earliest reported analyses by DMS was performed by a torsion 
pendulum. The work evaluated the effect of different conditionings on various 
commercial composites [91]. Samples stored at 37°C in water had lower modulus and 
larger damping than samples stored at 37°C under dry conditions. Running the same 
sample twice showed that additional cure resulted from analysis at elevated temperatures. 

Ferracane and Greener used DMS to examine the effects of the ratio of Bis-GMA 
to triethylene glycol dimethacrylate as well as the type of initiator system on 
thermomechanical properties [3]. This work demonstrated that the storage modulus at 1 1 
hertz was lower over the temperature range tested (25-150°C) in polymers with lower 
degree of conversion; i.e., that modulus increases with increasing degree of conversion. 
Samples with higher degrees of conversion had higher glass transition temperatures. This 
work also indicated that some additional cure takes place during DMS testing at elevated 
temperatures. 

Wilson and Turner have also performed DMS on dental polymers [92]. Their 
approach differs from that of other researchers in that Tg is reported at the significant loss 



43 

in storage modulus rather than where the tan 8 plot is maximized. Two methods of 
polymerization were used: y-radiation and photopolymerization. Bis-GMA and 
TEGDMA were the monomers used. The Tg values from the different interpretations 
and cures (Table 2.4) show that elastic modulus estimates are lower than the tan 8 
estimates. The photopolymerized samples are undercured as indicated by the Tg values. 
The modulus loss is related to a (3-relaxation, which has been identified in other studies 
on Bis-GMA [93] type polymers as well as lower molar mass ethylene glycol 
dimethacrylates [28, 30]. The P-relaxation in dimethacrylates is insensitive to different 
frequencies with the error of measurement [93]. The molecular origins of P-relaxations 
in dimethacrylates are not well understood. A more in-depth discussion of this 
phenomenon will be given later in this work. Other work has shown that the Tg of 
similar materials range between 85°C and 130°C [3]. Similar samples that were cured at 
80°C had glass transition temperatures that ranged from 151 to 195°C. The glass 
transition values obtained from the y-radiation polymerized samples are extremely high 
and may be indicative of the properties of a network cured to its physical maximum. It is 
interesting to note that the glass transition of fully cured diethylene glycol methacrylate 
has been estimated at 500°C [27]. This indicates that gamma radiation produces a more 
uniformly cured matrix, but it is still not possible to link every pendent group into the 
network structure in order to reach a theoretically fully cured structure. 

Wilson has examined the thermomechanical transitions in TEGDMA monomers 
cured with various dose levels (0.15-2.0 Mrad ) of y-radiation[94]. Four types of 
transitions that were identified: 1) associated with monomer in low dose samples (-60°C); 



195 


138 


- 


120 


48 


-25 


60 


-25 


120 


20 



44 



Table 2.4: Glass transition temperature estimates obtained by Wilson and Turner 

Sample Type Tan Delta Elastic Modulus 

Estimate Estimate* 

£C£ (°C) 

y-radiation cure of Bis-GMA 

y-radiation cure of TEGDMA 

Photopolymerization of Bis-GMA(75%) / 

TEGDMA(25%) in air 

Photopolymerization of Bis-GMA(75%) / 

TEGDMA(25%) in Nitrogen 

Second run of photopolymerization in Nitrogen 

*Values obtained with Autovibron in tensile mode at 1 1 Hz. and 2.5°C/min. 
Source: Wilson and Turner [92] 



2) associated with interactions between the oxyethylene groups (-10°C); 3) associated 
with additional polymerization of partially vitrified material because it occurs slightly 
above the curing temperature and a downturn in the length versus temperature curve is 
observed; and 4) associated with the glass transition temperature. At high dose levels 
transition 3, sometimes referred to as a P-relaxation, is no longer detectable. Others have 
detected this transition with dielectric relaxation techniques [24, 25]. 

Summary 
Many approaches have been taken to minimize or offset the shrinkage that occurs 
upon polymerization. Currently, a non-shrinking system has not been introduced that 
meets all the requirements for use in the mouth. Reduction of polymerization shrinkage 
has been obtained with molecules that have a larger molar mass than Bis-GMA. 
Ultimately these systems will only offset polymerization shrinkage and viscosity will 
become a limiting factor for large molar mass molecules. Liquid crystalline molecules 
show promise, but the toxicity of the current molecules and their ability to form a LC 
phase at lower temperatures are still issues. In order to have zero shrinkage or an 



45 

expanding system, ring open molecules must be employed. Unfortunately, these 
molecules are hampered by slow reaction kinetics, moisture sensitivity, and crystallinity, 
which limits solubility in base monomers. There is still a need for reduced shrinkage in 
polymer composites for dental applications. 

This work explored the use of viscosity modifying comonomers systems based on 
monomethacrylates, or multicomponent analogs of ethylene glycol dimethacrylate that 
reduced the polymerization shrinkage and improved or negligibly affected mechanical 
properties of Bis-MEPP-based polymers when compared to the standard comonomer 
TEGMDA. The monofunctional comonomers examined were cyclohexyl methacryalate 
(CHMA), 2-phenyloxyethyl methacrylate (PEMA), and tert-butylcyclohexyl 
methacrylate (t-BCHMA). The comonomers were selected because they have a range of 
viscosity modifying properties and their corresponding polymers have a range of glass 
transition temperatures. The relationship between initial viscosity and heterogeneity was 
explored by formulating monomers systems that cure to the same glass transition 
temperature but had different initial viscosities. The mechanical properties that were 
tested were tensile strength, fracture toughness, and modulus. The network structure of 
these polymers was evaluated using dynamic mechanical spectroscopy. These properties 
were examined at a range of comonomer molar concentrations. 

The heterogeneous network structure of dimethacrylates has been studied 
extensively but many questions still remain. Much of the structure and organization 
detail that occurs during cure still remains inaccessible to direct observation. The size 
and distribution of the low and high cross-link density phases, heterogeneity, has still not 
been determined. Knowledge of the size and distribution of the heterogeneity is 



46 

important because it affects the structure and performance of cross-linked dimethacrylate 
polymers. Knowledge of these structures and how they form will allow the tailoring of 
monomer selection and cure cycles to improve properties. Xenon NMR and 
TappingMode™ Atomic Force Microscopy (AFM) were used to determine the size and 
distribution of heterogeneous structures in dimethacrylates. Fractured and microtomed 
surfaces were examined at the scale of 500 to 1000 nm and regions of low and high 
cross-link density were elucidated using the topographic and phase imaging capabilities 
of TappingMode™ AFM. 

The efficiency of the incorporation of nadic methyl anhydride and maleic 
anhydride in methacrylate polymers was explored using the model compound 2- 
phenylethyl methacrylate. The polymers were examined with FTIR, DSC, and proton 
NMR to quantify the incorporation. 









CHAPTER 3 
ALTERNATE DILUENT SYSTEMS FOR BIS-GMA AND BIG-GMA ANALOGS 



Relevant Background 
The need for dilution of Bis-GMA and its analogs to optimize their properties has 
long been recognized. Bis-GMA, named for the starting products used to synthesize it 
(bis-phenol A and glycidyl methacrylate), is a highly viscous monomer (2980 Pa.s or 
29800 cp [1]). This high viscosity leads to early vitrification, which limits conversion 
during cure. This viscosity also limits filler loading. Vitrification is defined as the 
process in which the glass transition temperature of the reacting system reaches the 
reaction temperature. Polymerization becomes diffusion limited at the vitrification point. 
Vitrification at low conversions and other reaction mechanisms leads to heterogeneous 
systems composed of low molar mass or low cross-link density phases and high molar 
mass or highly cross-linked phases. It has been proposed that the low cross-link density 
phases behave as defects [34]. Therefore, if the amount of low molar mass material of 
the system can be reduced, the mechanical properties of the resin can be improved. Since 
the increase in viscosity during cure causes vitrification, it is logical that a system with a 
low initial viscosity will delay the onset of vitrification. This hypothesis is supported by 
the relationship between initial viscosity and degree of conversion seen in the literature 
[2-4]. Reactive diluents such as triethylene glycol dimethacrylate (TEGDMA) are 
effective at lower levels, but large amounts negatively affect matrix properties by 
increasing polymerization shrinkage and water sorption. Shrinkage results in poor 
marginal adaptation of composite restorations, which can lead to reformation of the 

47 



48 

cavity. Polymerization shrinkage also results in tensile stresses in matrix materials 
adjacent to the glass filler as well as intra-matrix stress. These stresses lower mechanical 
properties. The monomers studied in this work have a higher molar mass per 
methacrylate group than standard reactive diluents such as triethylene glycol 
dimethacrylate (TEGDMA). It has been observed that methacrylates with larger molar 
masses per methacrylate group have a lower theoretical shrinkage [51, 52]. Methacrylate 
monomers with cyclic pendant groups also have the ability to reduce the viscosity of Bis- 
GMA resins more effectively than TEGDMA [46-48]. This ability allows modifications 
in two ways: 1) lower amounts of diluent can be used, which results in less shrinkage; 2) 
resin mixtures with the same theoretical shrinkage that have a lower initial viscosity. 
Resin mixtures can also be designed that are a compromise between these two scenarios. 
If a material is failing through its weakest points then minimizing this structure should 
increase the strength of the material. The weakest point of the materials discussed in this 
work is the low cross-link density phase. 

The monomers selected will also provide information about the P-relaxation that 
is present in dimethacrylates. The onset of P-relaxations seen in dimethacrylates in this 
work occurs at approximately 50°C. This relaxation is near oral temperature; therefore, 
understanding the molecular structures that cause it is important. Some work indicates 
that p-relaxations in dimethacrylate polymers are due to pendant methacrylates groups 
[26, 93]. Other studies attribute the relaxation to the precursors of larger scale 
cooperative motion of the glass transition and the beginnings of vitrification [95]. In 
radiation-cured dimethacrylate networks the P-transition has been identified as an artifact 
of cure temperature due to vitrification and trapped free radicals [94]. Although there is 
a p-relaxation associated with the pendant groups in methacrylates, their incorporation 



49 

may provide insight into the P-relaxation in dimethacrylates. The incorporation of 
monofuctional monomers should reduce the amount of partially reacted dimethacrylates. 
The magnitude of p-relaxation is affected by the size of the pendant group [96]. 
Substitution of methacrylates with smaller pendant groups for that of partially reacted 
dimethacrylates (pendant) should provide information about the P-relaxation. It is also 
interesting to note that the temperature of the P-relaxation in observed dimethacrylates is 
not sensitive to frequency, while the P-relaxation in methacrylates is sensitive to 
frequency. This suggests that the molecular mechanism of P-relaxations is different in 
dimethacrylate and methacrylates. 

The monomers selected for this study take advantage of concepts developed in 
previous studies. Monofunctional methacrylates with cyclic structures in the pendant 
group can reduce shrinkage while maintaining mechanical properties [47, 48]. In this 
study monomers with higher molar masses than previous monomers and variable glass 
transition temperatures will be studied. These monomers are also more hydrophobic than 
previous monomer systems. These systems will be light-cured, while, samples in 
previous studies were heat-cured at 70°C. Initial viscosities may be a greater factor in the 
final properties when resins are light-cured at room temperature. The monomers (the Tg 
corresponds to the polymers formed by the monomers) that were used are cyclohexyl 
methacrylate (CHMA) (Tg = 83°C), t-butylcyclohexyl methacrylate (t-BCHMA) (Tg - 
98°C), 2-phenyloxy ethyl methacrylate (PEMA)(Tg=54°C) (Figures 3:l-3.3);2,2'-bis-(4- 
methacryloyloxyethoxyphenyl) propane (Bis-MEPP) (Figure 2.4) was the base monomer. 

Given two monomer systems that form polymers with the same glass transition 
temperature, the monomer systems with a lower viscosity will vitrify later and allow 



50 




O 
O— CH 2 — CH 2 — O-C-C-CH3 



CH 



Figure 3.1: The structure of 2-phenyloxyethyl methacrylate (PEMA) 







o 



O C C CH, 



Figure 3.2: The structure of cyclohexyl methacrylate (CHMA) 



CH, 



CH, 



CH, 




II 



O C C CH 



CH 2 

Figure 3.3: The structure of t-butylcyclohexyl methacrylate (t-BCHMA) 



51 

higher conversions. Whether the higher conversion will result in a polymer with a less 
heterogeneous structure will be explored in this work. 

Polydisperse polymers have lower viscosities than their corresponding 
monodisperse systems [97]. A multicomponent diluent system with the same average 
molar mass as triethylene glycol dimethacrylate should result in a lower viscosity. The 
lower viscosity will allow increased conversion and improved mechanical properties. A 
multicomponent system may also allow the reduction of polymerization shrinkage 
through the use of a smaller amount of diluent to obtain an acceptable viscosity. The 
monomers used for the multicomponent diluent experiments were ethylene glycol 
dimethacrylate, diethylene glycol dimethacrylate, triethylene glycol dimethacrylate, and 
poly (ethylene glycol 400 g/mol, 600 g/mol, 1000 g/mol) dimethacrylate (Figure 3.4). 
Three mixtures will be created that have the same weight average molar mass as 
triethylene glycol dimethacrylate (TEGDMA). TEGDMA systems will also be 
formulated as a baseline. One mixture will contain all of the monomers mentioned. The 
second will contain all of the monomers except the 1000 molar mass monomer. The third 
will contain all of the monomers except for the 600 and 1000 molar mass monomers. 
The hypotheses that are tested in this work are: 

1 . Monofuctional methacrylates with a cyclic or aromatic pendant group can 
reduce polymerization shrinkage while improving or negligibly affecting 
mechanical properties when compared to a Bis-MEPP/TEGDMA baseline 
system. The wet modulus of a 65/35 weight percent ratio Bis- 
MEPP/TEGDMA system is 2.6 ± 0.1 GPa and wet fracture toughness is 
0.56 ±0.10 MPa.m 5 . An increase in modulus would be a modulus of 2.8 
GPa or greater and no change would be modulus values ranging from 2.5 



52 

to 2.7 GPa. Similarly, an increase in fracture toughness would be values 
of 0.75 MPa.m 05 or greater and no change would be fracture toughness 
values ranging from 0.50 to 0.74 MPa.m 5 . The percent shrinkage for the 
baseline Bis-MEPP/TEGDMA systems is 9.7 ± 0.4 and an improvement 
would be shrinkage values less than 9.0 percent. The viscosity of the 
baseline Bis-MEPP/TEGDMA systems would be 780 ± 20 centipoise an 
improvement would be values lower than 740 centipoise. 

2. Heterogeneity can be reduced in polymers designed with the same glass 
transition temperatures by having a lower viscosity monomer system. 

3. Blends of TEGDMA analogs are more effective at reducing the initial 
viscosity compared to TEGDMA alone. 

Materials and Methods 

All monomers diluted with acetone and passed over inhibitor removal columns 
obtained from Aldrich. The acetone was removed from the monomers by evaporation 
under vacuum. The t-BCHMA monomer was obtained with catalyst still present, which 
resulted in an orange color. The catalyst was titanium tetrabutoxide. The catalyst was 
removed by reacting it with water. The t-BCHMA monomer was extracted from the 
water/precipitate mixture with chloroform in a separation funnel. The chloroform was 
removed from the t-BCHMA monomer by evaporation under vacuum. 

The initiator system that will be used for all systems is 0.5 weight percent 
camphoroquinone and 0.5 weight percent N,N-dimethyl-p-toluidine. Samples were cured 
using a UniXS twin Xenon strobe lamp light cure oven manufactured by Heraeus Kulzer. 
The resins or pastes were clamped between glass plates and polyethylene terephthalate 



53 



Table 3.1: Materials used 



Materials 



Abbreviation Manufacturer 



2,2'-bis-(4- 


Bis-MEPP 


methacryloylethoxyphenyl) 




propane 




cyclohexyl methacrylate 


CHMA 


tertiary-butylcyclohexyl 


t-BCHMA 


methacrylate 




2-phenyloxyethyl methacrylate 


PEMA 


triethylene glycol dimethacrylate 


TEGDMA 


ethylene glycol dimethacrylate 


EGDMA, 


analogs up to 600 g/mol. 


DEGDMA, 




TetraEGDM, 




EG(400), 




EG(600) 


poly (1000) ethylene glycol 


EG(1000) 


dimethacrylate 




Camphoroquinone 


CQ 


N, N-dimethyl-para-toluidine 


DMPT 



Sartomer Co., Exton, PA, USA 



Sartomer Co., Exton, PA, USA 
Esstech, Essington, PA, USA 

Sartomer Co., Exton, PA, USA 
Sartomer Co., Exton, PA, USA 
Sartomer Co., Exton, PA, USA 



Polysciences, Warrington, PA, 

USA 

Aldrich Chemical Co., Milwaukee, 

WI, USA 

Aldrich Chemical Co., Milwaukee, 

WI,USA 



(PET) film with a piece of Tygon tubing as a spacer (Figure 3.5). The mold was then 
illuminated on both sides for 180 sec. The distance between the glass plates was checked 
at various points to make the plates parallel. This ensured uniform samples. DMS and 
flexure samples were then cut using a slow speed diamond saw. Tensile samples were 
made in a similar manner to the flexure and DMS samples except an aluminum mold 
made to ASTM type 5 dogbone specifications was clamped between the glass plates and 
PET film. All samples for mechanical properties and DMS testing in the wet state were 
prepared in the same manner. The cured samples were placed in ultrapure water (16-18 
MQ-cm) at 37°C for 1 month. The water was changed periodically to facilitate monomer 
leaching and to better simulate the oral environment. The dry samples were aged in a 
desiccator for 1 week after curing and then tested. The monofunctional 






54 




n=1 Ethylene glycol dimethacrylate 

n=2 Diethylene glycol dimethacrylate 

n=3 Tri" 

n=4 Tetra" 

n=9 Poly(400) ethylene glycol dimethacrlate 

n=14Poly(600) " 

n=23Poly(1000) " 



Figure 3.4: The basic structure of poly (ethylene glycol dimethacrylate) analogs 



55 

monomers were added at 30, 40, 50, and 60 mol.% to the Bis-MEPP base monomer. The 
multicomponent diluent systems were added to the Bis-MEPP base monomer at 35 
weight percent. The matched Tg systems were formulated to have a Tg of 135°C 
(denoted by the maximum in the tan 8 peak at 1 Hz) using data from the previous 
systems. Table 3.2 contains the corresponding weight fractions of the resin formulations. 



Table 3.2: Molar fractions, weight fractions, and viscosities of Bis-MEPP/CHMA, Bis- 
MEPP/t-BCHMA, and Bis-MEPP/PEMA resins 



Resins 


Molar fractions 


Mass Fractions 


Viscosity (cP) 


Bis-MEPP/CHMA 


70/30 


87.7/12.3 


2170 ±10 




60/40 


82.1/17.9 


1330 ±10 




50/50 


75.3/24.7 


880 ±10 




40/60 


67.1/32.9 


460 ± 10 


Bis-MEPP/PEMA 


70/30 


83.7/16.3 


2030 ±10 




60/40 


76.7/23.3 


1390 ±10 




50/50 


68.7/31.3 


890 ±10 




40/60 


59.4/40.6 


590 ± 10 


Bis-MEPP/t-BCHMA 


70/30 


80.6/19.4 


2640 ±10 




60/40 


72.7/27.3 


1640 ±10 




50/50 


64.0/36.0 


990 ±10 




40/60 


54.3/45.7 


600 ± 10 



Flexure testing, for the measurement of the flexural modulus, was performed 
according to ASTM D790-81 three-point bend test procedure. An Instron 1 122 fitted 
with an environmental chamber operating at 37°C was used. The three-point bend test 
fixture has a support span of 40 mm and load span of 20 mm. The cross-head speed used 
is 1 mm/min. The sample size was 2.5 mm x 2.7 mm x 45 mm. 

The tensile breaking strength was measured according to ASTM standard D638- 
89 on the Instron 1 122 fitted with an environmental chamber set at 37 ± 2°C. The 
samples, which were type five geometry, were held with pneumatic grips with a clamping 



56 

force of 70 psi (0.48 MPa). The cross-head speed was 0.05 inches/min (1.3 mm/min) 
corresponding with an elongation rate of 5 percent/min. 

Flaws resulting from the manufacture of samples often influence strength in 
mechanical properties tests. To remove noise that results from sample manufacture and 
isolate materials properties, fracture toughness was measured using quantitative 
fractography. Fracture toughness is typically measured using methods such as single- 
edge notched, short rod, compact tension, or double torsion. These methods require 
particular sample preparation techniques and are often labor intensive to implement. 
Fracture toughness can be measured from the fracture surface of an ASTM dogbone 
when the fracture stress is known. The following equation was used to determine 
fracture toughness for brittle failures with edge flaws [98]: 

0.5 i 0.5 



K lc =l.24a f {(aby 3 } 



f 

Where: Cf = fracture stress 

a = depth of crack 

b = half-width of crack 
This equation is based on the assumption that failure-initiating crack can be modeled as 
an equivalent semicircular crack. The following equation was applied if yielding behavior 
was observed [98]: 

(K Ic f ={\2n{G f ) 2 c}l[f -{0.2l2((7 f ) 2 /(a ys ) 2 }] 



57 



Tubing 



Resins or 
Pastes 



Glass 
Plate 



PET 
Film 



Figure 3.5: Schematic of mold used for sample manufacture. Source: [99] 









58 

Where: c = (ab)° 5 

(j) =1.57 

CT ys = yield stress 
This equation is based on the assumption that the radius of the plastic zone is small 
compared to the crack size and that the plastic deformation is occurring at the crack tip. 
These assumptions allow the application of linear elastic fracture mechanics. Fracture 
toughness was measured in both wet and dry samples. Similar equations were applied if 
corner flaws were observed: 



K IC ={lA2) 2 ±<j f ^ 



The flaw sizes were determined using an Olympus optical microscope at 100 and 
200 magnification. The depth and the half-width of the fracture initiating flaws were 
measured using a reticule. The magnitude of the length was calibrated using a standard 
glass scale at each magnification. 

Dynamic Mechanical Spectroscopy was performed on one sample from each 
system. DMS imposes a sinusoidal strain on a sample over a range of temperatures and 
frequencies. It provides information about relaxation processes in polymers such as glass 
transitions and sub-glass transitions. The information is collected in the form of storage 
modulus (E'), loss modulus (E"), and the loss dispersion (tan 6) as a function of 
temperature and frequency. A Seiko DMS1 10 (double cantilever type) interfaced with a 
Seiko SDM5600H Rheostation was used to analyze the thermomechanical response in 
flexure mode. Measurements were taken at frequencies of 0.1, 0.5, 1, 5, and 10 Hz over 
the temperature range, -140°C to 200°C at a heating rate of 0.75°C/min in a nitrogen 
environment. The sample size used was approximately 2.5 mm x 10 mm x 40 mm. 



59 

Fourier Transform Infrared Spectroscopy was used to determine the degree of 
conversion in the various network polymer systems. The monomer samples were 
prepared by placing two drops of material between two sodium chloride (NaCl) crystals. 
The cured material samples were prepared for analysis by cryo-milling. The powders 
were mixed with IR grade KBr powder and pressed into thin pellets. Backgrounds were 
taken before every run with a blank crystal or KBr pellet. The spectra obtained represent 
32 scans using a Nicolet 20SX FTIR. No baseline corrections were performed on the 
spectra. The degree of conversion or percent-unreacted double bonds was determined by 
comparing the peak areas of the vinyl group at 1637 cm" 1 of the monomers and the cured 
polymers. Peak fitting software was used to draw baselines, deconvolute, and measure 
the areas of these peaks (PeakFit v4 by SPSS Inc.). The magnitude of these peaks is 
related to the path length of the radiation, the molar absorptivity, and the concentration of 
the chemical species. The aromatic peak at 1608 cm" 1 is used as an internal reference 
peak to normalize the vinyl peak intensities in different spectra. The normalization 
makes the peak areas a function of the concentration of chemical species. The intensity 
of peak at 1608 cm' that is associated with aromatic bonds does not change during 
polymerization. The following equation was used to calculate the percent of unreacted 
double bonds (UDB)[100]: 

UDB(%) = (bc/ad)*100 
Where: a = the vinyl absorbance peak at 1637 cm" 1 of the resins or pastes 

b = the aromatic absorbance at 1608 cm" 1 of the resins or pastes 
c = the vinyl absorbance at 1637 cm" 1 of the cured materials 
d ■ the aromatic absorbance at 1608 cm" 1 of the cured materials 



60 

It should be noted that this technique only samples vinyl bonds. It is possible for 
molecules to exist without vinyl bonds due to consumption by termination and 
disproportionation that are not incorporated into the network structure. Termination in 
the free radical polymerization methyl methacrylate at 60°C occurs 79 percent by 
disproportionation and 21 percent by combination [101]. 

The viscosities of the various monomer mixtures were measured at 25 °C with a 
Brookfield CAP2000 cone and plate viscometer. Shear rates ranging from 2667 to 6667s" 
were used to determine shear rate dependence of the viscosity. The two cone types used 
are: cone #1, which has a 0.45° slope angle and a radius 1.51 1 cm, and cone #2, which 
has a 0.45° slope angle and a radius of 1 .200 cm. 

The shrinkage of the monomer systems was calculated from differences in density 
between the monomers and cured polymer systems. The monomer densities were 
obtained by the average of 5 measurements from 25 ± 0.03 ml volumetric flasks. The 
densities of the cured samples were measured 5 times using the Mettler 33360 Density 
Determination Apparatus (ASTM D792-86). 

Statistical analysis was performed using the Sigma Stat™ software package. 
Student's t-tests were used to statically compare the different sample sets. 

Results and Discussion 
Physical Properties 

The viscosity of the Bis-MEPP/monofunctional diluent systems decreases, as 

expected, with increasing concentration of diluent monomers (Figure 3.6). At the 30 

percent molar concentration the CHMA systems have a higher viscosity than the PEMA 

and t-BCHMA systems. At the higher mol.% of diluent, the viscosity of the Bis- 



61 

MEPP/CHMA is lower than the viscosity of the Bis-MEPP/PEMA systems. This 
indicates that differences in weight percent of diluent are a larger issue at lower 
concentrations (Table 3.2), which holds true for the t-BCHMA systems as well. 

The monofunctional diluent will be compared to TEGDMA on a weight percent 
basis because of the large difference in weight fractions that would result from comparing 
resins formulated on a molar basis. The multicomponent diluent systems will be denoted 
with the following abbreviations: EG (400), EG (600), EG (1000). The abbreviations 
represent the multicomponent diluent systems EGDMA (400), EDGMA(600) and 
EGDMA(IOOO) described earlier in this work. A comparison of the diluent systems at 35 
weight percent shows that CHMA is the most effective followed by PEMA diluent 
(Figure 3.7). This is due to its small molar mass. T-BCHMA is the only diluent that is 
not more effective on a per weight basis as indicated by comparison with TEGDMA in 
the Bis-MEEP system. Although the molar mass of t-BCHMA is smaller than that of 
TEGDMA its structure is more rigid and bulky as evidenced by higher viscosity of the 
Bis-MEPP/t-BCHMA systems. The ethylene glycol units in the backbone of the 
TEGDMA produce a more flexible structure, which lowers viscosity more effectively 
than stiff bulky structures. Isobornyl methacrylate, a high molar mass methacrylate with 
a bulky stiff ring pendant group, also reduced the viscosity less effectively than 
TEGDMA[48]. FTIR analysis indicated that t-BCHMA monomer contained 10.4 ± 6.1 
percent non-reactive contaminates, which depending on their exact nature may decrease 
the t-BCHMA's ability to lower viscosity. There is no significant trend toward lower 
viscosity in the multicomponent systems composed of molecules with a greater range of 
molar masses. It seems that molar mass dispersity might only have significant effects in 



3000 



2500 



2000 



o. 
o 



f 1500 

o 
o 

1000 



500 







62 



" 



Bis-MEPP/CHMA 
Bis-MEPP/PEMA 
Bis-MEPP/t-BCHMA 



ID 



70/30 60/40 50/50 

(Molar Ratio) 



40/60 



Figure 3.6: Viscosity measurements on Bis-MEPP/PEMA, Bis-MEPP/t-BCHMA, and 
Bis-MEPP/CHMA resins at 25°C 






63 



1000 



800 



3. 600 



o 

I 400 



200 









TEGDMA CHMA PEMA t-BCHMA EG(400) EG(600) EG(1000) 

35 Weight Percent Diluent 

Figure 3.7: Viscosity measurements of various diluents in the Bis-MEPP-based resin 
systems diluted at approximately 35 weight percent 





















64 

systems of greater molar mass. 

Methacrylates have been shown to have shrinkage of 22.5 ml/mol per double [52] 
bond upon complete conversion of the double bonds to single bonds (Tables 3.3-3.7). 
Decreasing the molar concentration of methacrylate groups present in the monomer 
system can reduce the shrinkage that occurs upon polymerization. One way to reduce the 
molar concentration is to increase the specific volume of the monomers or the molar mass 
of the monomers with respect to the methacrylate groups. The discrepancy between the 
theoretical shrinkage and the measured shrinkages is due to the incomplete reaction of 
these systems. 



Table 3.3: Predicted theoretical and measured shrinkage values for Bis-MEPP-based 
copolymer systems 



Co-Monomer 



Molar Mass of comonomer* (g/mol) 
# of moles of comonomer in 1ml of mixture 

#ofmolesofBis-MEPP 

# number of moles C=C from co-monomer in 

lml of mixture 

# number of moles C=C from Bis-MEPP in lml 

of mixture 

Total # number of moles C=C in lml of mixture 

Theoretical volumetric shrinkage (%) 

Measured volumetric shrinkage 



35 wt.%. 
TEGDMA 



33 wt.% 
CHMA 



31 wt.%. 
PEMA 



*The molar mass of Bis-MEPP is 452 g/mol. 



286 
1.35x10 
1.59x10" 



,-3 



v3 



6.38xl0" 3 
13.2 
9.7 



168 
2.07x1 0" 3 
1.58x10 



-3 



5.23x10 
11.8 
9.0 



-3 



206 
1.68X10 -3 
1.68xl0" 3 



2.70xlO" J 2.07xl0" 3 1.68xl0" 3 
3.58xl0" 3 4.14xl0" 3 3.36xl0" 3 



5.04x10" 
11.3 

8.7 



For example, the Bis-MEPP/t-BCHMA systems, which have the highest average specific 
volume per methacrylate group, display the least amount of shrinkage. The systems that 
shrink the most upon curing are the Bis-MEPP/EDGMA-type systems, which have the 
lowest specific volume per methacrylate group of all the systems examined in this study. 



65 

This corresponds with results reported in other work where monomers with larger 
specific volume per methacrylate groups displayed less polymerization shrinkage [47, 48, 
50-52]. 

The percent shrinkage values within each system increase with increasing amount 
of diluent. This is expected in Bis-MEPP/CHMA and Bis-MEPP/PEMA systems 
because the molar mass per methacrylate group is smaller than that of the Bis-MEPP. In 
Bis-MEPP/t-BCHMA systems shrinkage also increases with increasing amounts of 
diluent. 



Table 3.4: Physical properties of Bis-MEPP/PEMA systems 

Composition Monomer density Polymer density Shrinkage Water uptake 

(molar ratio) (g/cm 3 ) (g/cm 3 ) (%) (wt.%) 

70:30 1.1 13 ±0.001 1.203 ±0.001 8.1 0.5 ±0.1 

60:40 1.105 ±0.001 1.197 ±0.002 8.2 0.5 ±0.1 

50:50 1.100 ±0.001 1.196 ±0.002 8.7 0.3 ±0.1 

40:60 1.098 ±0.002 1.193 ±0.002 8.7 0.4 ±0.1 



The systems examined here are not taken to a high degree of conversion as FTIR 
experiments presented later in this work will show. The incomplete reaction results in 
smaller amounts of polymerization shrinkage than theoretically predicted. 

Water uptake is lower in all of the monomethacrylate systems than the EGDMA- 
based systems. This is due to the lower polarity of the monomethacrylate systems. There 
is no difference in the water uptake within the error of measurement in the Bis- 
MEPP/PEMA systems. This is due to the similar structure of PEMA to Bis-MEPP. The 
water uptake experiments in the Bis-MEPP/CHMA systems have similar results to the 



66 

Table 3.5: Physical properties of Bis-MEPP/CHMA systems 



Composition 


Monomer 


Polymer density 


Shrinkage 


Water uptake 


(molar ratio) 


density (g/cm 2 ) 


(g/cm 2 ) 


(%) 


(%) 


70:30 


1.097 ±0.002 


1.186 ±0.001 


8.0 


0.7 ±0.1 


60:40 


1.089 ±0.002 


1.180 ±0.001 


8.4 


0.6 ±0.1 


50:50 


1.078 ±0.002 


1.170 ±0.001 


8.6 


5 ±0.1 


40:60 


1.065 ±0.002 


1.161 ±0.001 


9.0 


0.6 ±0.1 



Bis-MEPP/PEMA systems. There is also not much difference in the polarity of the 
structure of Bis-MEPP and CHMA. It has been shown that water uptake decreases in 
methacrylate-dimethacrylate polymers as the amount of methacrylate feed is increased 
[73]. This difference in the results seen in these systems and those examined in this work 
lies in sample preparation. The samples in the cited study were heat-cured to high 
conversion and then extracted prior to water uptake experiments. The samples in this 
work are light-cured at room temperature. This results in a larger amount of pendant 
groups and regions of low cross-link density material. It has been postulated that areas of 
low cross-link density take up solvents preferentially [30]. This preferential uptake is the 
mediating factor in the water uptake experiments. The Bis-MEPP/t-BCHMA systems 
have lower water uptake than the other systems. This could be due to the less polar 
structure of the t-BCHMA monomer. The lower value could also be due to loss of 
material while the samples are immersed in water. Bis-MEPP/t-BCHMA polymers lost 
an average of 8 weight percent in extraction studies performed with chloroform while the 
other polymers in this study had no detectable weight loss. No conclusions about the 
trend toward lower water absorption with increasing concentrations of t-BCHMA are 
possible due to leaching of materials during the hydration process. 



67 



Table 3.6: Physical properties of Bis-MEPP/t-BCHMA systems 



Composition 


Monomer density 


Polymer density 


Shrinkage 


Water uptake 


(molar ratio) 


(g/cm 2 ) 


(g/cm 2 ) 


(%) 


(%) 


70:30 


1.08110.002 


1.159 ±0.002 


7.2 


0.3 ±0.1 


60:40 


1.068 ±0.002 


1.147 ±0.002 


7.4 


0.3 ±0.1 


50:50 


1.052 ±0.001 


1.133 ±0.002 


7.7 


0.2 ±0.1 


40:60 


1.036 ±0.001 


1.1 18 ±0.002 


7.9 


<0.1 



Table 3.7: Physical properties of Bis-MEPP/EGDMA-type resins 



Diluent system 


Monomer 


Polymer density 


Shrinkage 


Water uptake 




density (g/cm 2 ) 


(g/cm 2 ) 


(%) 


(%) 


TEGDMA 


1.097 ±0.001 


1.204 ±0.002 


9.7 


1.0 ±0.1 


EGDMA(400) 


1.098 ±0.002 


1.206 ±0.002 


9.7 


1.0 ±0.1 


EGDMA(600) 


1.097 ±0.002 


1.203 ±0.002 


9.6 


1.0±0.1 


EGDMA(IOOO) 


1.100 ±0.001 


1.205 ±0.001 


9.5 


1.0 ± 0.1 



The degree of conversion values obtained from the FTIR spectra correspond well 
with those seen in the literature (Tables 3.8 and 3.9) [3, 102-105]. Other studies using 
monomethacrylates have reported values for degree of conversion that were slightly 
lower when mixed at 54 mol.% with Bis-GMA resins [47, 48]. A true comparison is not 
possible between these systems and those used in this study because the Bis-GMA 
systems were heat-cured at 70°C for 8 hours in a nitrogen atmosphere. Cure under such 
conditions results in a different network structure because the elevated temperature 
delays vitrification and the nitrogen eliminates oxygen inhibition of the polymerization. 
The polymerization process is also different due to differences in the concentration of 
radicals when heat curing and light curing are used. The lower conversion in these 
systems is due to their high viscosity and the difference in the curing techniques 
mentioned earlier. The higher viscosity is a result of the hydroxyl groups on Bis-GMA 



68 

molecules and its slightly higher molar mass compared to Bis-MEPP. The viscosity of a 
Bis-GMA system diluted with 54 mol.% tetrahydrofurfuryl methacrylate was nearly 2000 
cP [48]. Bis-MEPP resin formulated with similar mol.% of diluents had viscosities lower 
than 1000 cP. 

The degree of conversion increases with higher concentrations of diluent (Table 
3.8). The three factors that control the degree of conversion are viscosity of the initial 
monomer formulation, the Tg of polymer formed, and the average functionality of the 
system. All three of these factors are changed in ways that promote higher degrees of 
conversion by adding diluent, i.e., viscosity is lowered, average functionality is 
decreased, and the Tg of polymer formed is also decreased. Table 3.9 contains the degree 
of conversion values for the Bis-MEPP/EGDMA-type resins and the corresponding 
monomethacrylates with approximately the same weight percent diluent. At the same 
weight percent the Bis-MEPP/CHMA and Bis-MEPP/PEMA systems have a smaller 
average number of remaining vinyl bonds; however, the actual weight percent of the 
PEMA in the system is 31. A Bis-MEPP/PEMA system formulated with 35 weight 
percent would have a smaller quantity of remaining double bonds. The Bis-MEPP/t- 
BCHMA systems average value is lower than that of the Bis-MEPP/EDGMA-type 
systems; however, the error in measurement indicates that there is not a significant 
difference. The lower degree of conversion of the Bis-MEPP/t-BCHMA systems is due 
to the high viscosity of the monomer systems and the high Tg of the polymer system 
formed. 



69 



Table 3.8: Degree of conversion values based on percent vinyl bonds remaining in Bis- 
MEPP /PEMA, Bis-MEPP/CHMA, and Bis-MEPP/t-BCHMA systems 



Molar Ratio 


Bis-MEPP/PEMA 


Bis-MEPP/CHMA 


Bis-MEPP/ 
t-BCHMA 


70/30 


33.6 ±3.9 


36.0 ± 4.7 


40.4 ± 2.6 


60/40 


23.7 ±2.5 


25.1 ±3.1 


31.8 ±3.3 


50/50 


22.5 ± 3.2 


24.3 ± 2.6 


24.7 ± 3.5 


40/60 


18.1 ±3.1 


21.3 ± 3.6 


22.3 ±2.6 



Table 3.9: Degree of conversion values based on percent vinyl bonds remaining in Bis- 
MEPP/EDGMA-type systems, and the Bis-MEPP/PEMA, Bis-MEPP/CHMA, and Bis- 
MEPP/t-B CHMA systems formulated to approximately 35 weight percent dilu ent 

System Percent remaining vinyl bonds 

Bis-MEPP/TEGDMA 27.5 ± 4.4 

Bis-MEPP/EGDMA(400) 25.3 ± 3.0 

Bis-MEPP/TEGDMA(600) 27.9 ±3.4 

Bis-MEPP/TEGDMA( 1 000) 27.0 ± 4.2 

Bis-MEPP/PEMA 22.5 ±3.2 

Bis-MEPP/CHMA 21.3 ±3.6 

Bis-MEPP/t-BCHMA 24.7 ±3.5 



The mechanical properties data is presented as a comparison of different 
comonomers grouped by mol.%. The Bis-MEPP/PEMA, Bis-MEPP/CHMA, and Bis- 
MEPP/t-BCHMA systems will also be compared to Bis-MEPP/TEGDMA-type systems 
with similar weight percent formulation. The mechanical properties of each system are 
grouped by the molar ratios of the monomer components of the polymers (Figures 3.8- 
3.10). 
Flexure Testing 

The flexure modulus values from the Bis-MEPP/PEMA, Bis-MEPP/t-BCHMA, 
and Bis-MEPP/CHMA polymers are grouped by mol.% formulation (Figure 3.8). There 
are no pervasive trends with the change in mol.% of diluents in the dry or wet state. 



70 

The average modulus of the Bis-MEPP/PEMA resins varies 0.2 GPa throughout 
the entire range of PEMA concentrations in both the wet and dry states. Student's t-tests 
found no significant difference (P > 0.05) in any of the groupings in the Bis- 
MEPP/PEMA series. 

There are also no clear trends in the modulus of dry Bis-MEPP/CHMA systems. The 
60/40 system seems to be anomalously low (P < 0.05). In the wet Bis-MEPP/CHMA set 
there is a significant increase in modulus (P < 0.05) as CHMA is added up to 50 mol.%. 
Further addition of CHMA past 50 mol.% does not result in an increase in modulus upon 
hydration. The 50/50 and 40/60 systems are not significantly different (P < 0.05). The 
increase in modulus corresponds with a decrease in water uptake. Examination of the 
dynamic mechanical spectra of the 70/30 Bis-MEPP/CHMA systems shows that the glass 
transition is lower than the 60/40 Bis-MEPP/CHMA system, while the magnitude of the 
tan 8 maximum in the 70/30 system is less than the 60/40 system. This indicates that 
there is a large amount of low cross-link density material and pendant groups. The 
regions of low cross-link density and the increased water uptake result in the decreased 
modulus compared to the systems with larger amounts of diluent. Further discussion of 
the DMS is presented later in this work. The average values of modulus in the wet state 
were not significantly different from the dry state for all of the resins accept the 70/30 
composition. The modulus of the 70/30 composition is significantly (P < 0.05) decreased 
in the wet condition. 

There are no clear trends in Bis-MEPP/t-BCHMA systems in the wet or dry 
states. Pair- wise comparison of all of the dry compositions indicates that there is not a 
significant difference (P > 0.05). In the wet condition, the 60/40 composition is 



71 



CD 
0_ 
O 

CO 

_D 









PEMA-Dry 

PEMA-Wet 

CHMA-Dry 

CHMA-Wet 

t-BCHMA-Dry 

t-BHCMA-Wet 






70/30 



60/40 



50/50 



(Molar Ratios) 



I 1 



40/60 



Figure 3.8: Flexure modulus values of wet and dry Bis-MEPP/CHMA, Bis- 
MEPP/PEMA, and Bis-MEPP/t-BCHMA polymers at 37°C. The legend represents the 
monomers used to diluent the Bis-MEPP monomer 



72 

significantly greater (P < 0.05) than the other compositions. Pair-wise comparison of the 
same systems in the wet and dry state shows that the wet modulus is significantly greater 
(P < 0.05) in the 60/40 composition. 

Pair-wise comparison of the different polymer systems by mol.% formulation in 
the dry state indicates that the Bis-MEPP/t-BCHMA systems have significantly (P < 
0.05) lower modulus than the Bis-MEPP/CHMA and Bis-MEPP/PEMA systems in all 
cases except the 60/40 systems where the Bis-MEPP/PEMA is not significantly different 
from the Bis-MEPP/t-BCHMA systems. There is not a significant (P > 0.05) difference 
between the Bis-MEPP/PEMA and Bis-MEPP/CHMA systems in any of the 
formulations. The addition of monomethacrylates results in an increase in the molar 
mass between cross-links as evidenced in the DMS plots present later in this work. The 
different methacrylates result in different glass transition temperatures even though the 
average functionality of the mol.% formulations are the same. The tan 8 curves in the 
DMS have a shoulder on the low temperature side of the main relaxation. Correlation of 
the shoulder of tan 8 curves with the E' plots shows the E' values drop concurrently with 
shoulder. It is thought that the shoulder in the tan 8 curve, which is associated with a P- 
relaxation, is responsible for the similarity of the flexure modulus values. A more in- 
depth discussion of P-relaxations will be given later in this work. 

The results of a pair-wise statistical comparison of the different polymer systems 
by mol.% formulation in the wet state are shown in Table 3.10. The Bis-MEPP/t- 
BCHMA systems have the lowest modulus values with the exception of the 60/40 mol.% 
system. Generally speaking the Bis-MEPP/t-BCHMA polymers have the highest glass 
transition temperature and their monomer systems have the highest viscosities. The 



73 




presence of contaminates also results in a reduction 


in the modulus of the Bis-MEPP/t- 


BCHMA systems. 




Table 3.10: Pair-wise comparison of the modulus values of different polymer systems by 


mol.% formulation in the wet state using Student's t-test 


Molar Ratio Significantly different 


Not Significantly Different 


(P < 0.05) 


(P > 0.05) 


70/30 CHMA/PEMA, PEMA/t-BCHMA 


CHMA/t-BCHMA 


60/40 


PEMA/t-BCHMA, CHMA/PEMA, 




CHMA/t-BCHMA 


50/50 CHMA/t-BCHMA, PEMA/t- 


CHMA/PEMA 


BCHMA 




40/60 CHMA/t-BCHMA, PEMA/t- 


CHMA/PEMA 


BCHMA 




In the 70/30, 50/50, and 40/60 Bis-MEPP/t-BCHMA systems the modulus values are not 


significantly different (P > 0.05). There is a smaller amount of water uptake in these 


systems than in the 70/30 and 60/40 systems, but the water uptake values are suspect due 



to the possible loss of materials during the water treatment; therefore, differences in water 
uptake cannot be cited for the lack of modulus increase in the 70/30, 50/50, and 40/60 
systems. It is likely that there is some limit at which the loss of extractable material no 
longer results in increased modulus. The 70/30, 50/50, and 40/60 systems have reached 
this limit with a final weight loss of 6.5, 8.5, and 1 1.0 percent, respectively, after 
extraction in chloroform. 

The small range of modulus values observed in this work can be understood by 
examination of the dynamic mechanical spectrum of these polymers (Figures 3.1 1-3.21). 
There is a shoulder present on the low temperature side of the main relaxations in all the 
tan 6 spectra. The samples were tested at 37°C, which is in close proximity to onset of 



74 

this relaxation. This relaxation is likely in great part due to the incomplete curing of 
these systems as well as P-relaxations common to the methacrylate family. Further 
discussion about the nature of p-relaxations will be given later in this work. The 
incomplete cure of the systems results in a large of amount of heterogeneity, regions of 
low and high cross-link density materials. The systems with a higher average 
functionality form more highly cross-linked polymers; however, they form a less uniform 
network structure with a great amounts of defective structures. This structure and the 
incomplete cure of the samples results in small variations in the modulus values. A small 
range of modulus variation (0.4 GPa) has also been reported in other monomethacrylates 
that were incorporated into Bis-GMA systems [47, 48]. In order to confirm that the low 
temperature shoulder is a contributing factor of the modulus values a 70/30 Bis- 
MEPP/PEMA sample was post cured at 140°C for 2 hours immediately after the initial 
light curing. The post cure resulted in a dramatic reduction in the shoulder associated 
with the P-relaxation and a slight increase in the glass transition temperature as well as an 
increase in the cross-link density (Figure 3.24). The modulus values at 37°C tested using 
the Instron and procedures described earlier were 3.0 ± 0.1 and 2.7 ± 0.1 GPa for the post 
cured and normally cured sample, respectively. Although, the P-relaxation is greatly 
reduced by the post cure at 140°C, it has not been completely removed. This indicates 
that there is an underlying mechanism for this relaxation as well as the one that is an 
artifact of cure temperature. 

Understanding the results of flexure testing in the wet state as well as the changes 
in the modulus upon hydration requires knowledge of the processes that affect network 
structure during hydration. Hydration can increase modulus, decrease modulus, or have 



75 

no effect on modulus depending on the amount of water absorbed and how long the 
samples are soaked in water [5, 21, 23, 41]. The mechanisms responsible for these 
phenomena are plasticization by water and unreacted material, the leaching of materials 
from the system and physical and chemical aging. Swelling in water has also been shown 
to increase conversion [106, 107]. Dynamic mechanical spectroscopy results presented 
later in this work also indicate that additional polymerization or aging is occurring during 
hydration. Additional polymerization might also be occurring independently of water 
absorption. Diacrylates can continue to polymerize in the glassy state after the initial 
curing provided that the system has not vitrified or gelled to the level that diffusion is no 
longer reasonable [108]. The dry samples were tested one week after polymerization and 
the wet samples were tested one month after polymerization. The additional aging time 
of the wet samples may allow additional polymerization and physical aging to occur that 
would result in increased modulus values. In the Bis-MEPP/CHMA and Bis- 
MEPP/PEMA systems it is unlikely that leaching of materials is taking place in 
appreciable amounts because weight loss was not detected in extraction experiments 
performed in acetone and chloroform. The increase in the average modulus values at 
37°C, although not significant in most cases, from the wet to dry state is a result of 
additional polymerization and physical aging. The effect water has on the additional 
polymerization and physical aging is difficult to discern. The concentration of water is 
small and may aid the previously mentioned processes or it might result in a loss of 
modulus due to plasticization. Testing samples dry aged from 1 month might give more 
insight into the role the water is playing. 

In the Bis-MEPP/t-BCHMA system the 60/40 composition is notable because 
there is a significant increase (P < 0.05) in modulus in the wet state. A 5 percent weight 



76 

loss was observed in chloroform extraction of the 60/40 Bis-MEPP/t-BCHMA system. 
The loss of unreacted material coupled with the additional polymerization and physical 
aging are responsible for the observed results. 
Fracture Toughness and Tensile Testing 

The tensile strength data and fracture toughness data will be discussed 
simultaneously in this section. Caution should be used when interpreting tensile strength 
data because of the sensitivity of tensile strength to flaws induced by sample preparation. 
This is why fractographic analysis was used to interpret the strength properties of these 
materials. Tensile strength is an indicator of materials performance; however, it is not a 
materials property. It should also be noted that there was greater difficulty obtaining wet 
samples in which the flaws could be resolved than in the dry samples. 

The tensile strength of dry Bis-MEPP/PEMA polymers decreases significantly 
with the addition of PEMA (P < 0.05) (Figure 3.9). The 40/60 dry systems displayed a 
slight yielding behavior. The average yield stress of the 40/60 set was 41+2 MPa 
slightly higher than the average break stress, 38 ± 2 MPa. The yielding behavior is a 
result of the large molar mass between cross-links and the low glass transition 
temperature of the system as indicated by DMS (Figure 3.1 1). Fracture toughness 
(Figure 3.10) shows an increasing trend with the addition of PEMA; however, it is not 
significant (P > 0.05). This indicates that the tensile strength values may be a result of 
artifacts from sample preparation. There is not a significant difference (P > 0.05) in wet 
tensile strengths with the exception of the 70/30 system, which is significantly stronger (P 
< 0.05) than the other systems. There is not a significant difference (P > 0.05) between 
fracture toughness values in the wet state with the exception of the 40/60 system being 
significantly tougher (P < 0.05) than the 70/30 system. Again, the fracture toughness 



77 

results are different than tensile strength results. The difference in the trends is due to 
flaws resulting from sample preparation. The similarity of the strength values of this 
system can be attributed to the same factors that resulted in small variations in modulus 
for the Bis-MEPP/PEMA systems. As discussed earlier there is a P-relaxation in the 
dynamic mechanical spectra of these polymers that correlates with a decrease in modulus 
values. The p-relaxation is stronger in the samples with higher cross-link densities. The 
P-relaxation has been associated with a loose defective interlayer [34, 94] and regions of 
higher mobility [26] that are a result of vitrification. Although the cross-link density is 
higher in systems with less PEMA, it seems that the networks formed are less 
homogeneous than the networks formed when more PEMA is present. The less 
homogeneous network structure is indicated by the presence of the shoulder in the 
systems with less PEMA and also the larger full-width half maximums of the tan 8 peaks 
(Figure 3.1 1 and Table 3.1 1). The samples with higher concentrations of PEMA have a 
more uniform network structure. The heterogeneity of the higher cross-link density 
systems results in moderation of the strength properties of the systems; hence, the similar 
values of fracture toughness and tensile strength for all of the systems. The difference in 
the fracture toughness of the 70/30 and 40/60 Bis-MEPP/PEMA systems in the wet state 
may be a result of the weakening of the 70/30 systems due to preferential swelling in the 
regions of lower cross-link density. This mechanism has been proposed for the loss in 
fracture toughness of tetraethylene glycol dimethacrylates swelling in water [30]. 
Comparison of the wet and dry samples of the same mol.% formulations indicates that all 
the wet polymers have significantly higher (P < 0.05) tensile strengths than their dry 
counterparts. Again the fracture toughness values contrast the tensile strength values. 
There is no significant difference (P > 0.05) between fracture toughness of the wet and 



78 

dry Bis-MEPP/PEMA samples. The wet state of the 70/30 system appears to have quite a 
bit lower fracture toughness than the dry state, but the scatter on the measurement is 
large. The mechanisms discussed previously in this work that affect the modulus of these 
systems when they are hydrated also affect their strength properties. The additional 
polymerization and relief of residual stress that occurs over time in these samples is 
influenced by the presence of water. It is difficult to say if the presence of water 
enhances the addition polymerization and relief of residual stress or if it results in a 
reduction in properties compared to a system aged for a similar amount of time without 
water present. 

The 50/50 and 60/40 wet Bis-MEPP/CHMA systems have significantly (P < 0.05) 
higher tensile strengths than the corresponding dry systems (Figure 3.9). There is no 
significant difference between tensile strengths of the wet and dry state of the 70/30 and 
the 40/60 systems (P < 0.05). There is no significant difference (P > 0.05) in the tensile 
strength values of the different mol.% formulations within the wet and dry states of the 
Bis-MEPP/CHMA system. There is no significant difference (P < 0.05) in the fracture 
toughness values for any of the Bis-MEPP/CHMA systems wet or dry or by pair- wise 
comparison of wet and dry systems with the same mol.% formulation. The same 
arguments presented for the Bis-MEPP/PEMA systems can be applied to the Bis- 
MEPP/CHMA systems. Examination of the DMS plots of these systems shows a 
shoulder located on the low temperature side of tan 5 peak in the higher cross-link 
density systems. The shoulder is not as pronounced in the shoulder in the Bis- 
MEPP/PEMA systems but its onset is near the same temperature. The lack of resolution 
is likely due to the increased width of the main portion of the tan 8 in the Bis- 
MEPP/CHMA systems (Table 3.11). 



79 

The tensile strength in the Bis-MEPP/t-BCHMA systems decreases with 
increasing concentration in the dry state with the exception of the 70/30 formulation. In 
the wet state there is a decreasing trend with the addition of t-BCHMA. Table 3.9 
contains the results of statistics comparing the formulations in the wet state, the 
formulations in the dry state, and the pair- wise comparison of the formulations in the wet 
and dry state. The fracture toughness behavior is similar to the tensile strength behavior: 
additions of t-BCHMA greater than 40 mol.% result in a decrease in strength in the wet 
and dry state. The decrease is not significant between any of the systems in the 
dry state (P > 0.05). The 70/30 and 60/40 Bis-MEPP/t-BCHMA systems are 
significantly stronger (P < 0.05) than the 40/60 systems in the wet state. Pair-wise 
comparison of the wet and dry states of systems with the same mol.% formulations 



Table 3.1 1: Student's t-tests comparing tensile strengths of the various Bis-MEPP/t- 
BCHMA formulations in the wet and dry state and the same formulation in its wet and 

dr y state 

Significantly different Not Significantly Different 

(P < 0.05) (P > 0.05) 

70/30w vs. 40/60w, 60/40d vs. 40/60d, 70/30w vs. 70/30d, 60/40d vs. 50/50d, 
70/30d vs. 40/60d, 50/50d vs. 40/60d 70/30w vs. 50/50w, 70/30w vs. 60/40w, 
40/60w vs. 40/60d 60/40d vs. 70/30d, 60/40d vs. 40/60w, 

60/40w vs. 60/40d, 60/40w vs. 40/60w, 
60/40w vs. 50/50w, 50/50w vs. 40/60w, 
50/50w vs. 50/50d 



indicates that there is not a significant (P > 0.05) difference in fracture toughness between 
any of the systems. As mentioned earlier, extraction studies resulted in weight loss in the 
Bis-MEPP/t-BCHMA systems and FTIR studies indicated that approximately 10.4 ± 6.1 
percent of the vinyl bonds were reacted in the t-BCHMA monomer. The DMS spectra of 



80 

the Bis-MEPP/t-BCHMA systems indicate similar network structure to the Bis- 
MEPP/PEMA and Bis-MEPP/CHMA systems, the shoulder on the low temperature side 
of the tan 5 peak. The high viscosity and Tg of the systems with low concentrations t- 
BCHMA result in early vitrification and a heterogeneous network structure. As 
mentioned earlier, additional polymerization and the relief of residual stress of the 
samples results in improvement of the fracture toughness and tensile strength. This 
mechanism is more prominent in samples with lower concentrations of t-BCHMA. The 
water uptake in these samples is also low; therefore, little reduction in properties results 
from its presence. In these samples, leaching of unreacted materials during hydration 
can also be considered as a mechanism that will improve properties. In the 70/30 and 
60/40 Bis-MEPP/t-BCHMA systems the concentrations of unreacted materials are small 
and its removal results in improvement in mechanical properties along with the factors 
mentioned previously in the Bis-MEPP/CHMA and Bis-MEPP/PEMA systems. In the 
50/50 and 40/60 Bis-MEPP/t-BCHMA systems the leaching of unreacted material is 
great enough that the beneficial results of removing the material are minimized. As 
mentioned earlier, water uptake values are difficult to interpret because of the possible 
weight loss during hydration. The large amount of unreacted materials in the 50/50 and 
40/60 Bis-MEPP/t-BCHMA systems results in the reduction of their mechanical 
properties when compared to the 70/30 and 60/40 Bis-MEPP/t-BCHMA systems as well 
as lower average values when the wet state is compared to the dry state. 

The tensile strength and fracture toughness properties in the dry state are 
significantly lower (P < 0.05) in the Bis-MEPP/t-BCHMA systems than either the Bis- 
MEPP/CHMA or the Bis-MEPP/PEMA systems with the exception of the fracture 
toughness in 60/40 Bis-MEPP/CHMA and Bis-MEPP/t-BCHMA systems, which was not 



81 

significantly different (P > 0.05). The lack of significance in this system is due to the 
large scatter in the Bis-MEPP/CHMA systems and the small number of samples in the 
Bis-MEPP/t-BCHMA system. The low toughness and strength values of the Bis- 
MEPP/t-BCHMA systems can be attributed to presences of low molar mass materials due 
to contaminates as discussed earlier, as well as the high viscosity of the monomer 
systems and the high Tg of the polymer. The high viscosities and Tg result in early 
vitrification. The cross-link density is also lower in the Bis-MEPP/t-BCHMA systems 
than the Bis-MEPP/PEMA or Bis-MEPP/t-BCHMA, as indicated by the height of the tan 
8 peaks. Although the factors such as viscosity, Tg, and cross-link density all play a roll 
in the properties of the Bis-MEPP/t-BCHMA systems, their true properties are masked by 
the presence of contaminants and it is possible the difference in the strength properties is 
purely a result of the contaminants. There is a significant difference between the 40/60 
Bis-MEPP/CHMA and Bis-MEPP/PEMA systems (P < 0.05). The decrease in fracture 
toughness in the Bis-MEPP/CHMA systems is due to its more heterogeneous network 
structure. The Tg, as indicated by DMS, of the 40/60 Bis-MEPP/CHMA systems is 
higher than that of the Bis-MEPP/PEMA system by 20°C. It is likely that these factors as 
well as other differences in microstructures formed during polymerization result in a 
more heterogeneous network structure. This is indicated by the larger width of the tan 5 
peakofCHMA. 

The comparison of the wet state of different systems by formulations shows that 
Bis-MEPP/t-BCHMA systems have significantly (P < 0.05) lower tensile strengths than 
Bis-MEPP/CHMA and Bis-MEPP/PEMA systems. There is not a significant difference 
(P > 0.05) in the tensile strength values in any of the formulations of the Bis- 
MEPP/PEMA and Bis-MEPP/CHMA systems. Table 3.11 contains the results of the 



82 

statistical comparison of the fracture toughness values of the different systems in the wet 
state. The lack of significant difference in the wet state is partially due to the large scatter 
in the fracture toughness values and the small number of samples in some of the sets. 
The low fracture toughness in 50/50 and 40/60 Bis-MEPP/t-BCHMA systems can be 
attributed to the high viscosity of the monomers systems, Tg of the polymer formed, low 
cross-link density, and contaminates as discussed earlier. The presence of low molar 
mass materials is confirmed by the high amounts of extractable materials in these 
systems. In the 70/30 and 60/40 Bis-MEPP/t-BCHMA systems the leaching of the 
unreacted material resulted in some increase in the fracture toughness during the 
hydration period. Additional polymerization and the relief of residual stress are factors 



Table 3.12: Student's t-tests comparing wet fracture toughness values of the various 
formulations in the wet state 



Molar Ratio 


Significantly different 


Not Significantly Different 




(P < 0.05) 


(P > 0.05) 


70/30 




PEMA/t-BCHMA, CHMA/t- 






BCHMA, CHMA/PEMA 


60/40 


CHMA/t-BCHMA 


PEMA/t-BCHMA, CHMA/PEMA 


50/50 


CHMA/t-BCHMA, PEMA/t- 
BCHMA 


CHMA/PEMA 


40/60 


CHMA/t-BCHMA, PEMA/t- 
BCHMA 


CHMA/PEMA 



that may increase the fracture toughness. The similarity in the results of all the Bis- 
MEPP/PEMA and Bis-MEPP/CHMA polymers can be attributed to the heterogeneous 
network structures. The areas of low cross-link density are dictating the properties in 
both systems, which results in little difference in the tensile and fracture properties. 
Water uptake is not a factor because it is within the error of measurement for all of the 






83 

systems except the Bis-MEPP/t-BCHMA system, which is lower, but the values are 
suspect due to leaching of material during hydration. 
Dynamic Mechanical Spectroscopy 

Dynamic mechanical spectroscopy was used to determine the glass transition and 

P-transition temperatures of the network polymers as well as the change in modulus with 
temperature. Modulus information is important because of the range of temperatures 
experienced by dental restorations. The area of the tan delta peak is an indication of the 
heterogeneity of the networks formed. The width of tan delta peak is a measure of the 
distribution of environments seen by the polymer chains. The height of the tan delta peak 
is an indication of the average molar mass between cross-links. Table 3.13 contains a 
summary of the peaks statistics for the tan 8 peaks of all the dry polymers tested in this 
work. The glass transition temperatures are reported as the temperature at which the 
maximum in tan 8 occurs. As might be expected when comparing the mol.% 
formulations of the different systems, the Bis-MEPP/t-BCHMA systems generally have 
the highest glass transition temperatures and the Bis-MEPP/PEMA systems have the 
lowest. This can be related to the glass transition of the homopolymers formed by 
monomethacrylates. The homopolymer of t-BCHMA has a higher glass transition 
temperature due to its stiff bulky structure; therefore, the polymers it forms with Bis- 
MEPP have higher glass transitions temperature than the Bis-MEPP/PEMA systems. The 
homopolymer PEMA has lower glass transition temperature due to its more flexible 
structure; hence, the polymers formed with Bis-MEPP have lower glass transition 
temperature than either the Bis-MEPP/PEMA or the Bis-MEPP/CHMA polymers. 
Similar arguments can be made for the Bis-MEPP/CHMA polymers. The glass transition 



84 



CO 
Q. 



60 



50 



40 



g5 30 
c/) 

1 20 

CD 



10 







. 






i 


T 


- 


I i 


i 


i 






i 










■■ PEMA-Dry 
■M PEMA-Wet 
^m CHMA-Dry 

" I 1 CHMA-Wet 

^m t-BCHMA-Dry 
^m t-BCHMA-Wet 



70/30 60/40 50/50 

(Molar Ratios) 



40/60 



Figure 3.9: Tensile strength values of wet and dry Bis-MEPP/PEMA, Bis- 
MEPP/CHMA, and Bis-MEPP/t-BCHMA polymers at 37°C 












85 



b 



CL 



CO 
CO 
CD 

c 
.c 

D) 
O 



(1) 

i_ 

13 
-i— > 
O 
CO 






Jl 



1.0 

0.9 

0.8 

0.7 

0.6 

0.5 

0.4 

0.3 

0.2 

0.1 

0.0 

70/30 60/40 50/50 40/60 

(Molar Ratios) 

Figure 3.10: Fracture toughness values of wet and dry Bis-MEPP/PEMA, Bis- 
MEPP/CHMA, and Bis-MEPP/t-BCHMA polymers at 37°C 



PEMA-Dry 

PEMA-Wet 

CHMA-Dry 

CHMA-Wet 

t-BCHMA-Dry 

t-BCHMA-Wet 



■■mill 




86 

temperature of the Bis-MEPP/t-BCHMA polymers might also be slightly low due to the 
contamination discussed earlier. This would account for the similarity in the glass 
transition temperatures of the 40/60 systems in the Bis-MEPP/CHMA and Bis-MEPP/t- 
BCHMA systems. The glass transition temperatures of the Bis-MEPP/EDGMA-type 
polymers are higher than all of the Bis-MEPP/t-BCHMA, Bis-MEPP/CHMA, and Bis- 
MEPP/PEMA systems because of their low molar mass between cross-links as indicated 
by the tan 5 max values, which are half the magnitude of the highest cross-link density 
monomethacrylate systems. The maximums in tan 8 increase with increasing 
concentration of monomethacrylate. The values for the same mol.% formulation are 
similar in the Bis-MEPP/CHMA and Bis-MEPP/PEMA systems. The values for the Bis- 
MEPP/t-BCHMA systems are generally higher. This may be due to the larger viscosity 
and glass transition temperatures of the Bis-MEPP/t-BCHMA systems, resulting in early 
vitrification. The early vitrification would limit the amount of cross-links and result in 
larger max tan 5 values. The contamination in the Bis-MEPP/t-BCHMA systems also 
results in a reduction in the cross-link density of the systems. The full width at half max 
(FWHM), which is an indicator the heterogeneity of the network structure, decreases with 
increasing concentration of monomethacrylates. The FWHM was not reported for the 
Bis-MEPP/EDGMA systems because the main relaxation is not resolved enough from the 
P-relaxation to measure the values. The width of the peaks appears to be much broader 
than the monomethacrylate systems. The FWHM values of the Bis-MEPP/PEMA 
systems are generally the lowest as might be expected due to the low glass transition 
temperature and similar viscosities to the systems with the same mol.% formulations. 
The FWHM values of the Bis-MEPP/t-BCHMA polymers are larger than the 
corresponding Bis-MEPP/PEMA systems, which is expected based on the glass transition 



87 

temperatures and viscosities of the Bis-MEPP/t-BCHMA systems; however, the FHHM 
values are lower than the Bis-MEPP/CHMA systems. The FWHM values of the Bis- 
MEPP/CHMA system would be expected to be larger than the Bis-MEPP/PEMA 
systems, based on the fact the Bis-MEPP/CHMA systems form high Tg polymers at 
similar formulations. Based on the same argument it might be expected that the FWHM 
values of the Bis-MEPP/CHMA systems would be smaller than the corresponding Bis- 
MEPP/t-BCHMA systems, but they are not. The FWHM for the Bis-MEPP/CHMA 
polymers is exceptionally large and are likely a result of the specific details of network 
formation that will be discussed in-depth later in this work. 

As might be expected, the glass transition, as indicated by the maximum in the tan 
8, decreases with increasing amounts of PEMA in the Bis-MEPP/PEMA polymers 
(Figure 3.1 1). There is a corresponding decrease in the rubbery modulus values and 
increase in the magnitude of the tan 5 peak with increasing concentrations of PEMA, 
both of which correspond to a decrease in cross-link density. This is expected due to the 
reduction in the average functionality of the system and thus the cross-link density, as 
larger amounts of PEMA are added. There is a notable shoulder on the lower 
temperature side of the tan 8 peak. The shoulder is most prominent in the samples with 
lower concentrations of the PEMA. The elastic moduli (E') decrease concurrently with 
the shoulders. These shoulders have been attributed to a P-relaxation in dimethacrylate 
polymers. 

As mentioned earlier P-relaxations in dimethacrylates have been attributed to a 
number of different physical phenomena. Some have attributed P-relaxations in 
dimethacrylates to unreacted pendant groups [26]. p-relaxations in methacrylates have 



88 



Table 3.13: Statistics on tan 8 plots and glass transition activations energies for polymers 
in the dry state at 1 Hertz 



System 


F WHM (°C) 


Tan 8 Max 


Glass Transition 
Temperature (°C) 


Bis-MEPP/ 








PEMA 








70/30 


63 


0.29 


139 


60/40 


51 


0.34 


127 


50/50 


44 


0.39 


123 


40/60 


38 


0.47 


114 


Bis-MEPP/ 








CHMA 








70/30 


86 


0.28 


136 


60/40 


71 


0.31 


142 


50/50 


67 


0.37 


138 


40/60 


62 


0.41 


134 


Bis-MEPP/ 








t-BCHMA 








70/30 


65 


0.33 


149 


60/40 


55 


0.38 


148 


50/50 


54 


0.45 


142 


40/60 


51 


0.53 


135 


Bis-MEPP/ 
TEDGMA 


N/A 


0.16 


172 


Bis-MEPP/ 

EDGMA(400) 
Bis-MEPP/ 

EDGMA(600) 
Bis-MEPP/ 

EDGMA(IOOO) 


N/A 
N/A 
N/A 


0.14 
0.14 
0.12 


170 
167 
170 



been associated with pendant groups; however, p-relaxations in methacrylates are 
sensitive to frequency [96]. As mentioned earlier, p-relaxations in this work and in other 
work on [93] dimethacrylates are not sensitive to frequency. p-Relaxations in the 
ethylene glycol dimethacrylate-type polymers have been attributed to the beginnings of 
vitrification [24, 95] and larger scale cooperative motions and localized oxyethylene 
chain motions coupled to the acrylate chain, the precursors of larger scale cooperative 
motions responsible for glass transitions. P-relaxations occur in glassy ethylene glycol 



89 

dimethacrylates in the range of 20 to 30°C[30]. The P-relaxations studied in this work 
occur at approximately 80°C and the polymers do not have a true oxyethylene structure. 
The presence of the bis-phenol-A group may stiffen the oxyethylene chain and result in 
the p-relaxation occurring at a higher temperature. Clarke has also reported similar 
transitions in Bis-GMA-type polymers [93]. The a and p relaxations in the study by 
Clarke follow the same trends with frequency as those in this study. The only difference 
is the relaxations in Clarke's study occur at higher temperatures. The difference is due to 
variation in the sample preparation methods. The samples in the study by Clarke were 
heat-cured at 80°C for 3 hours. The samples in this study were prepared by light-initiated 
cure at room temperature. Clarke argued that these peaks could not be regarded as a 
relaxation process because they are independent of frequency. The mechanism proposed 
of cooperative motion and the beginnings of vitrification proposed earlier may account 
for the p-relaxations' insensitivity to frequency. Wilson has identified P-relaxations in 
triethylene glycol dimethacrylate in various stages of cure by y-radiation [94]. The 
reactions were detected at approximately 40°C and were accompanied by a decrease in 
sample length. The decrease in sample length was attributed to additional 
polymerization. The relaxations were attributed to partially vitrified materials that are an 
artifact of the curing temperature. Studies in which tetra ethylene glycol dimethacrylates 
were cured initially at 55°C and then post cured at 155°C have reported p-relaxations in 
the range of 20 to 30°C. This suggests that the p-relaxation is not entirely an artifact of 
the polymerization temperature. The onset of the P-relaxation occurs at virtually the 
same temperature in the Bis-MEPP/t-BCHMA, Bis-MEPP/PEMA, and Bis- 
MEPP/CHMA systems in this work. It is likely that the Bis-MEPP/t-BCHMA, Bis- 



90 

MEPP/PEMA, and Bis-MEPP/CHMA samples achieve similar temperatures during their 
cure cycle. This suggests that the p-relaxations observed in this work are related to the 
temperature obtained during the cure cycle of these polymers. Several other factors also 
suggest that these relaxations are related to partially vitrified materials: 1) the apparent 
insensitivity of the p-relaxation to the concentration of monomethacrylate, i.e., the lack of 
resolution when the temperature at which the a-transition is decreased by the addition of 
monomer. This suggests that as the Tg of the polymers is brought down near the 
temperature of the P-relaxation that the mechanism for the transition is same and 2) the 
dramatic change in the P-relaxation in the post cured samples (Figure 3.24) shows that in 
these systems this relaxation is in large part due to a loose defective structure that is an 
artifact of the cure. The small shoulder left after the post cure indicates there is a 
relaxation present that can be attributed to P-relaxations common to the methacrylate 
family. 

The tan 8 peaks indicate that the location of the glass transition shifts with 
frequency in the dry 70/30 mol.% Bis-MEPP/PEMA systems (Figure 3.12); however, the 
temperature that the P-relaxations occur at is not sensitive to frequency within the error of 
measurement. The magnitude of the P-relaxation decreases with increasing frequency, 
while the magnitude of the ot-relaxation increases with increasing frequency. This may 
be due to the coupling of the P-relaxation with the a-relaxation. Examination of the E" 
spectra confirm that the P-relaxation is not sensitive to frequency. 

The glass transition temperature, as indicated by the maximum in the tan 8 peak, 
is lower in the wet systems compared to their dry analogs (Figure 3.13). The hydration of 
the polymers result in a merging of the a and P relaxations (Figure 3.13). As discussed 



91 

earlier hydration and aging have been shown to increase conversion in dimethacrylates. 
Polymerization has also been shown to continue in the vitreous state. If the P-relaxation 
is associated with the beginnings of vitrification and hydration and aging result in 
additional polymerization then the temperature at which the P-relaxation occurs would 
increase during hydration. The main relaxation (a) is associated with a more fully 
vitrified state that additional polymerization is unlikely to occur in during hydration; 
hence, hydration results in a decrease in the temperature of the main relaxations due to 
plasticization. These changes result in the merger of the relaxation in the tan 8 plots. 
Many of the same features of the Bis-MEPP/PEMA systems are seen in the E' 
and tan 5 plots of the Bis-MEPP/CHMA polymers (Figure 3.14). One notable difference 
is in the 70/30 systems. The glass transition as indicated by the maximum in the tan 5 
peak is at a lower temperature than the 60/40 resin systems. This is due to the large 
amount of pendant groups and low cross-link density material plasticizing this system. 
This phenomenon is not present in the 70/30 Bis-MEPP/PEMA systems because of the 
larger weight percent of PEMA at the 70/30 molar ratio. The Tg of 70/30 Bis- 
MEPP/PEMA polymer is 139°C while the Tg of the 70/30 Bis-MEPP/CHMA is 136°C. 
The P-relaxations of the Bis-MEPP/CHMA polymers are much less pronounced than 
those of the Bis-MEPP/PEMA polymers. As discussed earlier the onset of the relaxation 
begins at the same temperature in both systems but the Bis-MEPP/PEMA systems have a 
slight plateau region. The magnitude of the tan 5 curves in the Bis-MEPP/CHMA 
polymers increase in a more continuous manner. The main relaxation blends more with 
the p-relaxation in Bis-MEPP/CHMA. This less resolved shoulder is largely due to the 



92 



1e+10 



1e+9 



CD 

LJJ 

8> 



1e+8 



1e+7 




• 70/30 

• 60/40 

• 50/50 

• 40/60 



i n huh '" hi»""" ll 



-100 



100 

Temperature (°C) 



200 



0.5 



- 0.4 



- 0.3 



to 
0.2 .§ 



0.1 



- 0.0 



Figure 3.1 1: Log E' and tan 5 response of dry Bis-MEPP/PEMA polymers at 1 hertz 












93 



2.5e+8 



2.0e+8 



1.5e+8 - 



w 1.0e+8 
Eu 



5.0e+7 



0.0 



-100 




100 

Temperature (°C) 



200 



0.35 
0.30 

- 0.25 

- 0.20 



0.15 § 



0.10 
0.05 
0.00 



Figure 3.12: Log E' and tan 5 response of dry 70/30 Bis-MEPP/PEMA polymers at 
various frequencies 



94 



1e+10 



1e+9 



£ 

8> 



1e+8 



1e+7 




• 70/30 
» 60/40 

• 50/50 

• 40/60 



mssss^^Wism 



-100 



""•. • • 




■ar o oo° 







O □ T ^ 



100 

Temperature (°C) 



200 



0.5 



- 0.4 



0.3 



to 
0.2 S 



- 0.1 



0.0 



Figure 3.13: Log E' and tan 6 response of wet Bis-MEPP/PEMA polymers at 1 hertz 









95 

broader width of the main relaxation peak. The onset of the a-relaxation occurs at a 
lower temperature in the Bis-MEPP/CHMA polymers than the Bis-MEPP/PEMA 
polymers. The half max widths of the Bis-MEPP/CHMA systems are generally wider 
than those of the Bis-MEPP/t-BCHMA and Bis-MEPP/CHMA systems. Differences in 
the glass transitions, monomer viscosity, and the specifics of network formation may 
account for the broadening. A more in-depth discussion of the peaks widths will be given 
later in this work. There is also a sub-Tg relaxation or y-relaxation present at 
approximately -80°C in the Bis-MEPP/CHMA polymers. This relaxation has been 
assigned to chair-to-chair configuration changes in the cyclohexane group in poly 
(cyclohexyl methacrylate) [109]. The increase in the magnitude of the relaxation with 
increasing concentration of CHMA corroborates this hypothesis. The sub-Tg relaxation 
is also seen in the E" plots shown in Figure 3.16. As seen with the Bis-MEPP/PEMA 
systems, the P-relaxations are not affected by frequency (Figure 3.16). The E" plots 
confirm that the temperature at which the (3-relaxations occur is not sensitive to 
frequency. It is also interesting to note that the maximums in the E" plots occur at the 
temperature of the P-relaxations. The merging of the shoulder associated with the (3- 
relaxation and the main peak can be attributed to the insensitivity of the P-relaxation to 
frequency compared to the a transition. 

The trends of the DMS spectra of the wet Bis-MEPP/CHMA polymers (Figure 
3.15) are similar to those of the dry Bis-MEPP/CHMA resins. The glass transition 
temperatures decrease with increasing amounts of CHMA. The glass transition 
temperatures are lower than those of the corresponding dry resins. The P-relaxations in 
both the Bis-MEPP/CHMA and Bis-MEPP/PEMA are not as prominent in the wet 



96 

systems as they are in the dry systems. Again the merging of the shoulder with the main 
peak can be attributed to the same phenomena that cause the merger of the shoulder and 
the main peak in the Bis-MEPP/PEMA systems. The sub-Tg relaxation is still present in 
wet Bis-MEPP/CHMA polymers. 

The dry Bis-MEPP/t-BCHMA spectra display the same general trends as both the 
Bis-MEPP/CHMA and the Bis-MEPP/t-BCHMA systems (Figure 3.17). The 
temperature at which the glass transition occurs decreases with increasing concentration 
of the t-BCHMA and the magnitude of the maximum associated with the glass transitions 
increases. The increase in the magnitude is attributed to a decrease in the cross-link 
density or a larger molar mass between cross-links. The beginnings of the rubbery 
plateau region of the E' also indicate an increase in the molar mass between cross-links 
with increasing amounts of t-BCHMA. The tan 8 plots have a shoulder on the low 
temperature side of the main peaks that is more prominent in the formulations with low 
concentrations of t-BCHMA. The shoulder is associated with a P-relaxation that can be 
attributed to the same network structures that caused the relaxations in the Bis- 
MEPP/PEMA and Bis-MEPP/CHMA polymers. There is also very little change in the 
Tg values of the 70/30 and 60/40 formulations. The small amount of change is due to 
pendant groups and contaminants plasticizing the systems. As seen with the Bis- 
MEPP/PEMA and Bis-MEPP/CHMA systems, the storage moduli (E') decrease 
correspondingly with the shoulders in the tan 5 peaks. Hydration of the Bis-MEPP/t- 
BCHMA (Figure 3.18) systems results in the same phenomena in the tan 6 peaks seen in 
the Bis-MEPP/PEMA and Bis-MEPP/CHMA systems, which can be attributed to 









97 



1e+10 



1e+9 



CD 
CL 

LLI 

O 



1e+8 



1e+7 




■100 



- 0.4 



- 0.3 



- 0.2 



- 0.1 



0.0 



to 

g 



100 

Temperature (C) 



200 



Figure 3.14: Log E' and tan 8 response of dry Bis-MEPP/CHMA polymers at 1 hertz 



98 



1e+10 



1e+9 



03 

Eu 

8" 



1e+8 - 



1e+7 




■100 



100 

Temperature (°C) 



200 



0.6 



0.5 



0.4 



- 0.3 



0.2 



0.1 



0.0 



«3 
C 
CD 



Figure 3.15: Log E' and tan 8 response of wet Bis-MEPP/CHMA polymers at 1 hertz. 



99 



CO 

■ 

LU 



2.5e+8 



2.0e+8 



1.5e+8 



1.0e+8 



5.0e+7 



0.0 



10 Hz 
5 Hz 
1Hz 
0.5 Hz 



0.1Hz 




-100 



100 

Temperature (°C) 



200 



0.35 
0.30 
0.25 
0.20 



CO 

0.15 S 



0.10 
0.05 
0.00 



Figure 3.16: Log E' and tan 8 response of dry 70/30 Bis-MEPP/CHMA polymers at 
various frequencies 






100 

additional polymerization associated with aging and hydration as well as plasticization of 
the structures associated with the main relaxations. Examination of the different 
frequencies of the DMS (not shown) plots showed behavior similar to the Bis- 
MEPP/CHMA and Bis-MEPP/PEMA systems. 

The DMS of the Bis-MEPP/EDGMA type systems in the dry state (Figure 3.19) 
shows Show many of the same features seen in the monomethacrylate systems: the main 
peaks, the shoulder and the decrease in storage modulus corresponding with the shoulder 
in the tan 8 peak. One notable difference is the relative size of the shoulders associated 
with the P-relaxations. The shoulder appears large when compared to the magnitude of 
the main peak; however its magnitude is similar to the shoulders in the other systems. 
The small magnitude of the main peak makes the shoulder appear larger. There is also a 
broad relaxation present at approximately -100°C. Relaxations in this temperature range 
have been attributed to the presence of moisture in the polymer (-90°C) and localized 
motions in the oxy-ethylene chains (-130°C) [95]. The relaxation is broad and covers the 
temperature range from -140 to -80°C in the dry state. The relaxation appears more 
prominent in the wet state, indicating that it may in fact be due to moisture. Other work 
on dehydrated Bis-MEPP/TEGDMA resins also show a broad low temperature relaxation 
[99] that occur in the same temperature range as those seen in the dry. It is likely that 
both relaxations due to moisture and oxy-ethylene chain mobility are occurring but 
cannot be resolved. There is also a slow increase in magnitude of the tan 5 from 
approximately -50°C to the onset of the P-relaxation. It is difficult to attribute increase 
to one physical phenomenon. It is an indication of the overall heterogeneity of the 



101 



1e+10 



1e+9 



ro 

ill 
g 5 



1e+8 



1e+7 




Kiwwmums 



-100 



0.6 



0.5 



0.4 



0.3 



0.2 



- 0.1 



0.0 



100 

Temperature (°C) 



200 



to 

c 

TO 



Figure 3.17: Log E' and tan 8 response of dry Bis-MEPP/t-BCHMA polymers at 1 hertz. 















102 



1e+10 



1e+9 



03 
UJ 

5 



1e+8 



1e+7 




70/30 
60/40 
50/50 
40/60 



WSSSSBBMlimwmiiL 



-150 -100 -50 







0.6 
0.5 
0.4 
0.3 
0.2 
0.1 
0.0 



WO 
5 



50 100 150 200 250 

Temperature (°C) 



Figure 3.18: Log E' and tan 8 response of wet Bis-MEPP/t-BCHMA polymers at 1 hertz 



103 

systems formed and distribution of mobility in the network. The height of the tan 6 
decreases from the TEGDMA to the EG(1000) series. This is due to the smaller N 
average molar mass between cross-links associated with the increasing range of molar 
mass units in the diluent mixture. Similar behavior to the monomethacrylate systems is 
seen in the wet state of the Bis-MEPP/EDGMA-systems (Figure 3.20). The main 
relaxation decreases in temperature and the P-relaxation increases in temperature. The p- 
relaxation is more resolved in these systems than the monomethacrylate systems because 
the main relaxation occurs at a higher temperature. The multifrequency (not shown) plots 
also show similar behavior to the monomethacrylate systems. 

Bis-MEPP/CHMA, Bis-MEPP/t-BCHMA, and Bis-MEPP/PEMA polymers were 
formulated to the same glass transition temperature to isolate the effect of the viscosity of 
the monomer systems (Figure 3.21). It is difficult to draw conclusions about the effect of 
monomer viscosity on heterogeneity when comparing systems that have different glass 
transition temperatures because vitrification is related to the Tg of the polymer system 
forming. A polymer system with a high Tg is likely to vitrify sooner than one with a low 
Tg. The viscosities were 1710 ± 20, 460 ± 10, and 600+10 cp for the Bis- 
MEPP/PEMA, Bis-MEPP/CHMA, and Bis-MEPP/t-BCHMA systems, respectively. The 
widths of the tan 5 peaks at half max were 59, 62, and 51°C for the Bis-MEPP/PEMA, 
Bis-MEPP/CHMA and Bis-MEPP/t-BCHMA systems, respectively. There appears to be 
no correlation between the viscosity of the monomer systems and the width of the tan 8 
peaks. The Bis-MEPP/CHMA system has the lowest viscosity of the formulated 
polymers yet it has the largest peaks width. The width of the Bis-MEPP/t-BCHMA 



104 



1e+10 



CO 

a. 



Qj 1e+9 
8> 



1e+8 





w A 


• TEGDMA 35 wt.% 

• EGDMA(400) 35 wt.% 
■ EGDMA(600) 35 wt.% 
o EGDMA(1 000) wt.% 




£ff • \%o 

V 



-100 



100 

Temperature (°C) 



0.18 
0.16 
0.14 
0.12 

0.10 eo 

c 
co 

0.08 K 
0.06 
r 0.04 
0.02 
0.00 



200 



Figure 3.19: Log E' and tan 6 response of dry Bis-MEPP/EDGMA type polymers at 1 
hertz 









105 



1e+10 



CD 
Q. 



uj 1e+9 



1e+8 




-100 



0.20 



0.15 



0.10 



(0 



- 0.05 



0.00 



100 

Temperature (°C) 



200 



Figure 3.20: Log E' and tan 8 response of wet Bis-MEPP/EDGMA type polymers at 1 
hertz 



106 



1e+10 



1e+9 



LU 
8> 



1e+8 



1e+7 




w tttmuum 



-100 



100 

Temperature (°C) 



200 



0.6 



0.5 



0.4 



0.3 



0.2 



0.1 



0.0 



c 

CO 



Figure 3.21 : Log E' and tan 6 response of dry Bis-MEPP/PEMA, Bis-MEPP/CHMA, 
and Bis-MEPP/t-BCHMA polymers formulated to a Tg of 135°C 



107 

system is narrower than the Bis-MEPP/PEMA systems. This is more in line with the 
expectation that a lower monomer viscosity and a larger molar mass between cross-links 
will result in a more homogeneous structure; hence, a narrow peak width. The origins of 
the large peak widths in the Bis-MEPP/CHMA system may be due to such factors as 
isomers of the CHMA, which are in turn due to its synthesis and specific structural 
sequences that form during polymerization, i.e., alternating type copolymers, linear 
segments composed of CHMA units, tacticity, etc. All of these structures are not 
detectable using any of the techniques in this work so their possible existence and affect 
on the DMS plots can only be speculated. The possibility of different isomers forming 
during the synthesis of these systems cannot be discounted; however, their effect would 
be expected to be pervasive in the all of the systems. Although the specific synthetic 
techniques of the CHMA and t-BCHMA monomers are not known, the techniques might 
be similar. Based on this similarity we might expect a similar response of the tan 5 
curves if the isomers were the cause of the broad line widths. Differences in tacticity 
might also result in a broader tan 8 peak. Again, we might expect similar results in the 
Bis-MEPP/t-BCHMA and Bis-MEPP/CHMA due to the similar structure; however, the 
effect of tacticity cannot be ruled out. Dielectric relaxation studies on polymers 
composed of TEGDMA and methyl acrylate suggest that linear segments of methyl 
acrylate are formed in the network structure [24]. Similarly, CHMA might form linear 
segments. CHMA is the smallest of the three monomers and may have the highest 
mobility, allowing it to diffuse through the partially reacted network and form linear 
chains of poly (CHMA). The PEMA and t-BCHMA monomers are larger than CHMA 
and may have more restricted diffusion through the partially reacted network, thus 
resulting in smaller regions of linear polymer. The lower viscosity of the Bis- 



108 

MEPP/CHMA systems when equal masses of the monomethacrylate diluent are 
compared suggests that CHMA has a higher mobility than PEMA or t-BCHMA; 
however, no specific data about the relative diffusivity of the these monomers in Bis- 
MEPP systems is available so these claims cannot be substantiated. Although no claims 
can be made that clearly explain the unexpectedly wide tan 8 peaks width seen in the Bis- 
MEPP/CHMA systems, the formation of linear segments of poly (CHMA) seems to be 
the most plausible. 

Master curves were made for the Bis-MEPP/PEMA and Bis-MEPP/CHMA 
systems (Figure 3.22) to obtain additional information about the network structure of 
these polymers (Table 3.14). The modulus of the Bis-MEPP/CHMA system decreases to 
smaller values at lower frequencies compared to the Bis-MEPP/PEMA systems. This is 
due to the lower cross-link density of the Bis-MEPP/CHMA system compared to the Bis- 
MEPP/PEMA system. The shift factors (Figure 3.23) of the Bis-MEPP/PEMA and Bis- 
MEPP/CHMA systems are similar in the range of the reference temperature (Table 3.14). 
The shift factors deviate at temperatures higher and lower than the reference temperature. 
At 60°C above the reference temperature the shift factors of the Bis-MEPP/CHMA and 
Bis-MEPP/PEMA systems differ by less than 1, while at 60°C below the temperature the 
factors differ by almost 5. The CI and C2 parameters of the WLF equation listed below 
can be related to the fractional free volume at Tg (f g ) and the 

-Q(r-r) 



log a t = 



C 2 +T-T g 



109 

thermal coefficient of expansion of the fraction free volume above Tg (a/). Ci is 
identified with B/2.303/^, and C2 is identified withy^/a/[l 10]. The values for Ci only 
vary slightly because the free volume of different polymers in the glassy state does not 
vary greatly. The smaller Ci value of Bis-MEPP/PEMA indicates that it has a greater 
free volume at Tg. The larger C2 value of the Bis-MEPP/PEMA system indicates that it 
has a smaller coefficient of thermal expansion. This smaller coefficient is due to its 
higher cross-link density. The master curves, tan 8, and shift factor plots give insight into 
the network structure beyond a simple measurement of the glass transition temperature. 
These results indicate that it is possible to formulate systems that have the same glass- 
transition temperature with different monomer systems. How the systems behave over a 
range of temperatures and frequencies is not purely a function of the glass transitions 
temperature. The details of the network formation such as cross-link density, specific 
chemical structures, and the distribution of mobilities in different phases dictate the 
response of the materials over a range of temperature and frequency. 

Table 3.14: WLF parameters used to obtain the master curve fits for the Bis- 
MEPP/PEMA and Bis-MEPP/CHMA systems formulated to a glass transition 
temperature of 135°C 



Polymer Ci C 2 Reference temperature Glass transition 

(T r ) temperature (T g ) 



Bis-MEPP/CHMA 30 175 60 80 

Bis-MEPP/PEMA 22 125 60 80 



110 

Comparison of Alternate Comonomer Systems by Weight to Triethylene Glycol 
Dimethacrylate 

Triethylene glycol dimethacrylate is considered a baseline or standard of 

comparison for comonomers in Bis-GMA type systems. This section will compare the 

mechanical properties of the monomethacrylate and multicomponent ethylene glycol 

dimethacrylate-based comonomer systems that were formulated at the weight percent 

ratio of approximately 65/35 Bis-MEPP/comonomer. The Bis-MEPP/EGDMA-type 

systems will be compared to themselves and the monofunctional monomers will be 

compared to TEGDMA in the wet state and the dry state. 

Flexure Testing 

There is not a significant difference between any of the systems based on 

EGDMA-type comonomer systems (P > 0.05) (Figure 3.25) except the EG(600) and 

EG(1000) systems in the wet state. This is due to the similarity in structure of all these 

monomer systems. It was originally hoped that formulating multicomponent diluent 

systems based on ethylene glycol units with a range of molar mass would lower viscosity 

more effectively than a TEGDMA system. The viscosity values were similar in all the 

multicomponent systems (Figure 3.6); hence, the properties of these systems are 

generally similar. There is a slight increase in cross-link densities within the increase in 

molar mass dispersity as indicated by DMS presented earlier in this work; however, the 

differences were not great enough to result in significant changes in mechanical 

properties. Similarly, the lack of difference between the wet and dry modulus can be 

attributed to the effects of aging in water. As mentioned earlier, aging samples in water 

has been shown to have little effect on, increase, or decrease modulus depending on the 

hydrophilicity of the systems in question and the hydration conditions [21, 41, 106]. The 



Ill 



1e+10 



1e+9 - 



LU 

9 



1e+8 



1e+7 



ie+10 



le+9 - 



LU 



1e+8 



Frequency vs Modulus of Bis-MEPP/PEMA 



i i i i i i i i i i i i i i i i i i l i l l l i l i i i i i i i i i i i i i r~r~i i i i i i i i 



Log Frequency (S 



1, 



••• ••• •< 



Frequency vs E' of Bis-MEPP/CHMA 



1e + 7 I i ■ i ' I I I I I I I I I I I I I I I I I I I I I i i i i i i i i i i i i i i i [ i | i i | 

Log Frequency (S" 1 ) 

Figure 3.22: Master curve plots of the Bis-MEPP/PEMA (top) and Bis-MEPP/CHMA 
(bottom) formulated to glass transition temperatures of 135°C (the frequency range of the 
x-axis is le-16 to le32 s" 1 .) 









112 



to - 

20 - 


• 

• 
■ 

■ 


• 
■ 










15 - 




• Bis-MEPP/PEMA 
■ Bis-MEPP/CHMA 




10 - 














t 


t 
• 


- 

-5 - 

-10 - 

-15 - 


i 


i 




• 
• 

■ 
1 1 1 1 1 I 



-80 -60 -40 -20 20 40 

T-TgCC) 



60 80 



100 



Figure 3.23: The plot of shift factors (a t ) calculated from the WLF equation for the Bis- 
MEPP/PEMA and Bis-MEPP/CHMA systems formulated to a Tg of 135°C 



113 



0.35 

0.30 

0.25 

0.20 

0.15 

0.10 

0.05 -\ 

0.00 



C 




• Post Cure at 140°C 
■ Normal Cure 



- 1e+9 




1e+10 



LU 



1e+8 



1e+7 



-100 



100 

Temperature (°C) 



200 



Figure 3.24: Log E' and tan 8 response of dry Bis-MEPP/PEMA with normal cure and 
post cure at 140°C Bis-MEPP/CHMA 



114 

competing mechanisms of additional polymerization and plasticization by water dictate 
the properties of the wet samples. In the case of the EGDMA-modified resins, the 
quantity of water absorbed is not great enough to cause a reduction in the average 
modulus through plasticization (P > 0.05). In the more hydrophobic monofunctional 
resins the plasticization caused by the water is not great enough to reduce the modulus of 
the materials and generally there is an increase in the average modulus. This theory is 
supported by the larger amount of water absorbed by the Bis-MEPP/EDGMA-type 
systems. 

There is not a significant difference between the Bis-MEPP/TEGDMA system 
and any of the monomethacrylates in the dry state (P > 0.05) except the Bis-MEPP/t- 
BCHMA system, which has a significantly lower (P < 0.05) modulus than all the other 
systems. The similarity in the modulus is surprising when the difference in cross-link 
densities is considered. The tan 8 peaks are extremely broad in the Bis-MEPP/TEGDMA 
systems and the presence of the (3-relaxations indicate a very heterogeneous network 
structure. The onset of the P-relaxation is at a lower temperature in the Bis- 
MEPP/TEGDMA systems than the monomethacrylate systems. As discussed earlier, the 
defective regions of the network and the incomplete cure of these systems, which is 
indicated by the shoulder in the tan 8 peaks, mediate the modulus values and result in 
little difference in modulus values despite larger differences in cross-link density. 

In the wet state the Bis-MEPP/CHMA system have significantly (P < 0.05) larger 
modulus values than Bis-MEPP/TEDGMA system. The modulus Bis-MEPP/t-BCHMA 
system is significantly (P < 0.05) lower than the other systems. The Bis-MEPP/PEMA 
and Bis-MEPP/CHMA systems are not significantly different (P > 0.05). The 
phenomena of plasticization by water and additional polymerization are responsible for 



115 

the higher modulus seen in the wet state of the Bis-MEPP/CHMA system. The 
difference seems to be in the amount of water absorbed by the Bis-MEPP/TEDGMA 
systems. The additional amount of water results in the reduction of modulus compared to 
the monomethacrylate systems, which absorbed less water. The low modulus values of 
the Bis-MEPP/t-BCHMA systems are due to the contamination present and may also be 
related to water uptake. It is difficult to make any conclusion about the properties of 
water uptake of the Bis-MEPP/t-BCHMA systems due to the possible leaching of 
material. 
Tensile Strength and Fracture Toughness 

The tensile strength (Figure 3.24) and fracture toughness (Figure 3.25) will be 
discussed simultaneously in this section. As might be expected from the flexure results, 
there is no difference (P > 0.05) in the fracture toughness or tensile strength for any of the 
Bis-MEPP/EDGMA-type systems in the wet or dry states. As discussed in the flexure 
testing results section, the differences in network structure are not great enough to result 
in changes in the mechanical properties. There is no significant difference (P > 0.05) in 
the dry tensile strength of the Bis-MEPP/t-BCHMA, Bis-MEPP/CHMA, and Bis- 
MEPP/TEDGMA polymers. The Bis-MEPP/t-BCHMA system has significantly lower 
tensile strength (P < 0.05) than all the other systems. The dry fracture toughness values 
of the Bis-MEPP/CHMA and Bis- MEPP/PEMA systems are significantly (P < 0.05) 
greater than the Bis-MEPP/t-BCHMA and Bis-MEPP/TEGDMA systems. There is no 
significant difference (P > 0.05) in the fracture toughness values of the Bis-MEPP/t- 
BCHMA and Bis-MEPP/TEDGMA systems. The differences in the tensile strength and 
the fracture toughness results can be attributed to the greater variability of the tensile 



2 - 



CD 
CL 

CO 

s 



116 





















] Dry 
I Wet 



TEGDMA EG(400) EG(600) EG(1000) PEMA CHMA t-BCHMA 

35 Weight Percent Diluent 

Figure 3.25: Flexure modulus values of wet and dry 65/35 weight percent Bis- 
MEPP/TEGDMA, Bis-MEPP/EGDMA multicomponent type, Bis-MEPP/CHMA, Bis- 
MEPP/PEMA, and Bis-MEPP/t-BCHMA polymers at 37°C. The legend represents the 
monomers used to dilute the Bis-MEPP monomer 


















117 

strength due to its flaw sensitivity. The higher fracture toughness values of the Bis- 
MEPP/CHMA and Bis-MEPP/PEMA is surprising, considering the lower cross-link 
density of these systems compared to the Bis-MEPP/TEGDMA systems. It has been 
suggested that strength values in cross-linked acrylate polymers are dictated by the loose 
defective phase or a low cross-link density phase [34]. The decrease in fracture 
toughness of the Bis-MEPP/TEDGMA despite its high cross-link density suggests that 
the fracture properties are being dictated by the low cross-link density phase. The DMS 
spectra presented earlier in this work suggests that the Bis-MEPP/TEDGMA system is 
more heterogeneous than the Bis-MEPP/CHMA and Bis-MEPP/PEMA systems. This 
heterogeneity can be associated with a more defective network structure; hence, lower 
strength. Similar arguments were made for the increased compressive strength when n- 
butyl methacrylate was added to tertramethacrylate-(bis-trimethylolpropane)-adipate 
systems[34]. 

There is no difference in tensile strength (P > 0.05) in the wet state of the Bis- 
MEPP/CHMA, Bis-MEPP/PEMA, and Bis-MEPP/TEDGMA. The Bis-MEPP/t- 
BCHMA system is significantly weaker than the other wet systems. The Bis- 
MEPP/CHMA and Bis-MEPP/PEMA systems have higher (P < 0.05) fracture toughness 
values than the Bis-MEPP/t-BCHMA and Bis-MEPP/TEDGMA. The Bis-MEPP/t- 
BCHMA and Bis-MEPP/TEDGMA systems do not have significantly different (P > 
0.05) fracture toughness values. The same claims for the higher fracture toughness 
values of the Bis-MEPP/CHMA and Bis-MEPP/PEMA systems in the dry state can be 
made for the wet state. The lower fracture toughness of the Bis-MEPP/t-BCHMA system 
is likely due to the network defects associated with the contamination present in the 



118 



DU 




1 i Ory 






■B Wet 


50 - 




-r 














'ro' 

1 40- 




-L 




£ 


I 


i- 




Strength 

CO 

o 

i 






-L 








fl 




Tensile 
o 


















10 - 
- 


I 






i 


i 









TEGDMA EG(400) EG(600) EG(1000) PEMA CHMA t-BCHMA 

35 Weight Percent Diluent 

Figure 3.26: Tensile strength values of wet and dry 65/35 weight percent Bis- 
MEPP/TEGDMA, Bis-MEPP/EGDMA multicomponent type, Bis-MEPP/CHMA, Bis- 
MEPP/PEMA, and Bis-MEPP/t-BCHMA polymers at 37°C. The legend represents the 
monomers used to dilute the Bis-MEPP monomer 









119 




TEGDMA EG(400) EG(600) EG(1000) CHMA PEMA t-BCHMA 

35 Weight Percent Diluent 

Figure 3.27: Fracture toughness values of wet and dry 65/35 weight percent Bis- 
MEPP/TEGDMA, Bis-MEPP/EGDMA multicomponent type, Bis-MEPP/CHMA, Bis- 
MEPP /PEMA, and Bis-MEPP/t-BCHMA polymers at 37°C. The legend represents the 
monomers used to dilute the Bis-MEPP monomer 



120 

t-BCHMA monomer. The fracture toughness of the pure system might be greater than 
the TEGDMA system, but it is unlikely that it would be much greater than the Bis- 
MEPP/PEMA and Bis-MEPP/CHMA systems due, to the high glass transition of the 
polymers it forms and the high viscosities of its monomers systems. The lack of 
difference in the fracture toughness of Bis-MEPP/CHMA and Bis-MEPP/PEMA, despite 
their difference cross-link density, suggests that there may be some threshold of cross- 
link density above which fracture toughness properties begin to degrade. The specifics of 
network structure such as cyclization, pendant groups, and inactive chain ends may also 
play a roll in dictating fracture properties. 

Conclusions and Future Work 
This work shows that monomethacrylates have potential as alternate comonomer 
systems for Bis-MEPP or Bis-GMA-type resin systems. PEMA and CHMA reduce 
viscosity more effectively than TEGDMA and yield polymers with lower quantities of 
remaining double bonds and less shrinkage. The lower viscosities will allow higher filler 
loadings; hence, composites with better properties. The higher fracture toughness values 
are also encouraging. The modulus values of these polymers are also comparable with 
those of Bis-MEPP/TEDGMA systems. The low water uptake of these systems is 
beneficial and may help reduce plasticization effects during long-term exposure to 
moisture. The low shrinkage values of Bis-MEPP/t-BCHMA systems are attractive; 
however, we cannot draw any conclusion about the mechanical properties due to the 
presence of contaminant. The higher viscosities of the Bis-MEPP/t-BCHMA monomer 
systems suggest they may not be effective as a viscosity modifying monomers. Again, it 
is difficult to say what the properties of the pure monomer would be. The Bis- 



121 

MEPP/CHMA and Bis-MEPP/PEMA systems have larger molar mass between cross- 
links than Bis-MEPP/TEDGMA systems. This may make them susceptible to creep in 
the oral environment. Long-term studies need to be performed to evaluate the properties 
of these systems and how they might change during long periods of time in the oral 
environment. Overall, these monomethacrylates are promising as alternate comonomers 
for Bis-GMA-type resins. 

The addition of the monomethacrylates results in a more homogeneous network 
structure, as determined by the width of the tan 8 peaks in DMS plots, compared to Bis- 
MEPP/TEGDMA systems. The P-relaxation in these systems is associated with a loose 
defective or low cross-link density phase (that is in large part due to the incomplete cure 
of these systems) strongly influences the properties of the polymers in this study. It 
results in normalization of the modulus values of all the systems in this work. When the 
monomethacrylates are compared to themselves there is very little difference in the 
strength properties despite differences in the glass transition temperature and cross-link 
density. The small amount of difference can also be attributed to low cross-link density 
phase of these materials dictating their strength properties. The more homogeneous 
network structure of the monomethacrylate-type systems correlates with improved 
fracture toughness values when they are compared to Bis-MEPP/TEDGMA systems. 
The similarity of the strength properties of the monomethacrylate type systems to 
themselves and their improvement over the Bis-MEPP/TEDGMA systems seems to 
indicate that there is some threshold of cross-link density above which strength properties 
begin to degrade. Studies with different cure cycles would elucidate whether there is an 
absolute threshold in dimethacrylates systems or if it is related to the conditions under 
which the polymers are formed. It would also be interesting to blend the 



122 

monomethacrylates with themselves as well as with TEGDMA to optimize the favorable 
properties of all the systems. 

The relationship of the viscosity of the monomer systems to heterogeneity of the 
network structure formed is not direct but viscosity is an indicator of the heterogeneity 
that might be expected. Other factors such as the glass transition temperature of the 
polymer formed, the average functionality of the systems and the details of the network 
formation strongly affect the heterogeneity of the polymer formed. It was also shown 
that polymer based on different monomer systems can be formulated to the same glass 
transition temperature. Although, the polymers have the same glass transition 
temperature their properties vary across a range of temperature and frequencies due to 
different distribution of mobility and structures and chemical differences in the systems. 
















CHAPTER 4 
CHARACTERIZING THE HETEROGENITY OF DIMETHACRYLATE POLYMERS 



Relevant Background 

It has been proposed that the weakest regions, those of low cross-link density, in 
the polymer determine the strength properties of dimethacrylates [34]. The 
heterogeneity of dimethacrylate polymers has been demonstrated experimentally using 
such techniques as dielectric spectroscopy [24, 25], dynamic mechanical spectroscopy 
[26-28], extraction and swelling experiments [29-31], photochromic [32], charge- 
recombination luminescence [33], paramagnetic probes [34], and CP/MAS NMR C 13 
spectroscopy [9, 35]. The heterogeneous network structure of dimethacrylates has been 
studied extensively but many questions still remain. Much of the structure and 
organization detail that occurs during cure still remains inaccessible to direct observation. 
The size and distribution of the low and high cross-link density phases, heterogeneity, has 
still not been determined. If the regions of low cross-link density are acting as flaws and 
failure is initiated in these regions then knowledge of the size of these regions may allow 
their correlation with fracture properties. This correlation may ultimately allow the 
tailoring of monomer selection and cure cycles for enhanced mechanical properties. 

The purpose of this work is to combine TappingMode™ AFM (TMAFM) and 
Xenon- 129 NMR methods to characterize the heterogeneity of dimethacrylate-type 
polymers as it develops during polymerization and in its final state. The heterogeneity of 
the fully cured polymers will be characterized using TMAFM with phase imaging 



123 



124 

capability. TMAFM uses a sinusoidal frequency imposed on the tip. When the tip 
interacts with the sample the frequency that it oscillates at deviates from the imposed 
frequency. The phase image is a representation or map of the difference between these 
frequencies. Different tip-sample interactions will result in different phase shifts for the 
regions of different cross-link density. 

The change in the NMR signal can provide information about how the different 
phases develop during cure [81]. 129 Xenon NMR was performed to further characterize 
the network structure and its formation. The xenon NMR signal is indicative of its 
chemical environment. The size of regions was estimated using Pulsed Field Gradient 
(PFG) NMR spectroscopy. The pulsed field gradient method allowed the calculation of 
the self-diffusion constant of xenon. With knowledge of the self-diffusion coefficient of 
xenon estimates were made of root mean square displacement in three dimensions via the 
following equations [86]: 



r(i) = j6Dt 



Where: "t" is the time over which the self-diffusion constant is measured. The self- 
diffusion constants "D" were calculated from the following relationship [90]: 

A(G) stands for the echo amplitude in the presence of a field gradient with the strength G, 
y is the gyromagnetic ratio of the considered nuclei, A is the time interval between the 
application of the field gradients and 8 is the time interval during which the gradient is 



125 

applied. By varying the strength of G, D can be obtained from the slope of the resulting 
curve. 

Xenon NMR studies were performed on dimethacrylate systems polymerized with 
an iniferter initiation system. Iniferter initiator systems were used to gain better control 
of the degree of conversion and to eliminate trapped radicals, which are detrimental to 
NMR signals. The iniferter initiation system is based on p-xylylene bis (N, N-diethyl 
dithiocarbamate) (XDT). Iniferters initiate polymerization when exposed to ultraviolet 
light. Unlike conventional initiators, iniferter systems form two types of radicals: a 
carbon radical and a sulfur-based dithiocarbamyl (DTC) radical [111]. The carbon 
radicals are reactive and capable of initiating polymerization, while the DTC radicals are 
less reactive and don't participate significantly in initiating polymerization. When the 
curing light is shut off, the DTC initiator fragments remain mobile and are able to 
combine with radicals remaining in the system, thus stopping polymerization. This 
prevents radical trapping and also halts polymerization in a more controlled fashion than 
a typical light-cured system. Another unique feature of XDT initiated systems is the 
ability of the DTC fragments to form "marcoiniferters" [111]. The DTC terminated 
polymer chains decay into radicals upon exposure to light and can reinitiate 
polymerization. Iniferter based systems are a valuable tool for studying network 
formation in dimethacrylate systems; however, the polymerization kinetics of XDT are 
slightly different from those of a conventional photoinitiator. This difference lies in the 
absence of the autoacceleration effect. Autoacceleration is due to an increase in the 
concentration of radicals because carbon-carbon radical termination becomes limited by 
diffusion. As previously mentioned, the DTC radicals remain mobile and thus are able to 



126 

terminate carbon radicals; hence, the concentration of radicals does not increase and 
autoaccelaration is not observed. Despite the differences, iniferter initiated systems form 
highly cross-linked heterogeneous networks that are analogous to systems polymerized 
with standard initiators [27]. 

Materials and Methods. 

The xenon nmr samples were composed of 2,2'-bis-(4 - 
methacryloyloxyethoxyphenyl) propane Bis-MEPP monomer. The iniferter initiator 
system was 1 .0 weight percent p-xylylene bis (N, N -diethyl dithiocarbamate) (XDT). 
The samples were created by placing the Bis-MEPP (XDT) solutions in 13 mm NMR 
tubes. The tubes were then placed on a vacuum line and evacuated at a 10" torr. 
Freeze-pump-thaw cycles were performed to aid in the removal of trapped air. The 
samples were then back-filled with xenon gas. A sufficient amount of gas to create a 
xenon pressure of approximately 10 atmospheres was then condensed into the tube using 
liquid nitrogen. While the xenon was condensed the tubes were flame sealed. The 
samples were then placed in a metal container where the xenon was allowed to sublimate. 
The samples were left in storage for 3 days to test the integrity of the flame seal and 
allow the xenon gas time to diffuse into the sample. The samples were polymerized in 
NMR tubes in the presence of xenon gas by exposing the samples to UV light for 10, 15, 
20, and 30 seconds, and 5 minutes. The samples were irradiated with a UVP Model 
UVL-56 black ray lamp. The long wavelength on the black lamp is 366 nm. 

Atomic force microscopy samples were prepared in two ways: fracturing and 
microtoming. Fractured samples were prepared from Bis-MEPP/ triethylene glycol 
dimethacrylate (TEGDMA) (60/40 wt%) polymers. The 70/30 Bis-MEPP/PEMA 



127 

samples, prepared as described in Chapter 3, were microtomed with an LKB/Bromma 
Ultratome to obtain a smooth surface for imaging. Bis-MEPP/TEGDMA samples were 
also microtomed. The post cured 70/30 Bis-MEPP/PEMA sample was prepared by 
placing it in an oven at 140°C for two hours immediately after the initial light curing. 
Bis-MEPP/TEGDMA samples were soaked in acetone for a period of 36 hours and then 
dried under vacuum to etch the sample to elucidate the morphology. 

Atomic force microscopy was performed using a Digital Instruments NanoScope 
III in the Tapping Mode™. The phase and height image were recorded for each sample 
examined. The height images are a representation of the topography of the sample 
surface. The phase image is representation of the difference between the free oscillating 
frequency of the AFM tip and the phase of the tip as it interacts with the sample. A more 
in-depth discussion of sample-tip interactions will be provided in the results and 
discussion section. Table 4. 1 is a compilation of the feed back parameters that were used 
to perform the scans. The scan sizes were 500 nm and 1000 nm or 1 micron. Plane fitting 
and flattening were performed on all images using the Digital Instruments software. 

Fourier Transform Infrared spectroscopy (FTIR) was performed after NMR was 
run on the samples. The samples were removed from their tubes and FTIR was then used 



Table 4. 1 : Feedback parameters for Tapping Mode™ Atomic Force Microscopy 

Parameter Range of adj ustment 

Integral Gain 0.1 -0.6 V 

Proportion Gain 0.2 - 1 .2 V 

Tuning Set Point 0.4 - 2.0 V 

Drive Frequency 300 ± 50KHz 

Drive amplitude 1 00 - 700 mV 



128 

to measure the conversion of vinyl bonds. Caution was used to avoid exposure to 
ambient light because it could initiate polymerization in the samples. FTIR spectra were 
obtained using a Nicolet 20SXB FT-IR spectrometer. A Harrick attenuated total 
reflection (ATR) stage was used with a 30° KRS-5 parallelogram crystal. The spectra are 
a result of 64 scans with an instrument resolution of 4 cm" 1 . The 5-minute cure sample 
was cryo-milled and then made into a KBr pellet. The pellet was measured in 
transmission. The quantity of vinyl bonds remaining or the degree of conversion was 
estimated using the equation given in Chapter 3. 

Xenon NMR was performed using a Bruker Avance 400 Digital NMR 
spectrometer equipped with a triple resonance 10 mm high-resolution probe. The 
magnetic field strength used was 9.4 T/m. A Bruker imaging gradient stack was used to 
apply the field gradients. All spectra are referenced to a xenon thermal standard located 
at a frequency of 1 10.7 MHz. All experiments were performed at 25°C. 

Standard xenon NMR spectra were collected from the samples prior to the 
diffusion measurements to locate the center frequency of the xenon signal in each sample. 
The parameters used to collect the spectra are located in Table 4. 2. Pulsed gradient 



Table 4.2: Parameters used to collect standard xenon spectra of Bis-MEPP(XDT) 
samples 



Parameter 


Value 


Sweep Width 


26455 Hz 


Acquisition Time 


0.4533 s 


Dwell Time 


18.9 ns 


Relaxation Delay 


40 s 


Pulse Length 


30 us 


Pulse Strength 


3.0 dB 


Number of FIDs Collected 


16 



129 

experiments were performed using stimulated echo and Hahn echo sequences. The pulse 
schemes for a standard or Hahn echo pulsed field gradient and stimulated echo pulsed 
field gradient sequence are pictured in Figure 4.1. The linearity of the field gradient 
strengths was calibrated by measuring the self-diffusion coefficients of water. The 
calibration experiments were performed on de-ionized water doped with 5 uM 
manganese chloride salt. The Mn salt is paramagnetic and enhances Tl relaxation times 
from 5 to ~0.2 seconds. The first experiments were performed using a standard pulsed 
gradient sequence with the parameters described in Table 4.3. The experiments were 
performed with 64 gradient steps. The measured self-diffusion coefficient for water was 
determined to be 2.10 ± 0.05* 10" 5 cm 2 /s. 



Table 4.3: Hahn echo sequence parameters used for water self-diffusion coefficient 

measurements 

Parameter Values 



90° Pulse 29 usee at 3 dB 

180° Pulse 29psecat-3dB 

5 - Gradient Time 5 msec 

G - Gradient Strength 5-900 mT/m 

A - Time Between Gradient Pulses 20 msec 



The self-diffusion coefficient of water was also measured using the stimulated echo 
sequence. The parameters used in the stimulated sequence were the same as the Hahn 
sequence with the exception of the absence of 180° pulse. In addition to the gradient 
strength range from 5 to 900 mT/m range experiments were also performed in the range 
10 to 210 mT/m. The sequences were performed with 16 and 64 gradient steps. The 
diffusion values from the 16 step and 64 step experiments performed in 10 to 210mT/m 



130 



Standard Pulsed Field Gradient Sequence 



Ml 


« 


71 






1 


G 


1 




1 


< 


T 


— H H 


x — 


> 



Stimulated Pulsed Field Gradient Sequence 

71/2 tc/2 K /2 



l 




J 



L 




-2x- 



> ^ 



Figure 4.1 : Pulse schemes of Hahn echo pulsed field gradient and stimulated echo pulsed 
field gradient sequence 









131 



gradient range were 2.18 ± 0.18*10' 5 cm 2 /s and 2.04±0.13*10" 5 cm 2 /s, respectively. The 
diffusion values from the 16 step and 64 step experiments performed in 5 to 900mT/m 
gradient range were 2.27 ± 0.12*10" 5 cm 2 /s and 2.15 ± 0.06*10" 5 cm 2 /s. Values reported 
from other pulsed gradient diffusion studies are 2.34 ± 0.08* 10" 5 cm 2 /s [90]. The same 
authors reported a value of 2.23 ± 0.05* 10" 5 cm 2 /s with steady gradient experiments. 
Simpson and Carr reported a value of 2.13 ± 0.15*10" cm /s [ 1 1 2] . 

Measurements were made on the Bis-MEPP(XDT) samples using the stimulated 
echo sequence. The parameters used in pulse sequence to measure the self-diffusion 
coefficient of xenon in the Bis-MEPP(XDT) systems are located in Table 4.4. 



Table 4.4: Parameters used in pulse sequence to measure the self-diffusion coefficient of 
xenon in Bis-MEPP(XDT) 



Parameter 



0,10, 15, and 20 
Second cure 



30 
Second cure 



90° Pulse 

Delay Time Between Scans 

5 - Gradient Time 

G - Gradient Strength 

A - Time Between Gradient Pulses 

Number of Gradient Steps 

Number of FIDs Collected 



3dB 

40 s 

5 ms 

5-700 mT/m 

0.5 s 

8 

64 



3dB 

40 s 

0.5 ms 

5-700 mT/m 

1.0 s 

6 
256 



Results and Discussion 
Atomic Force Microscopy 

Assigning phase and height components to atomic force microscopy (AFM) 
images is not a trivial task. The height and phase images depend on experimental 
parameters such as the free oscillation amplitude of the tip (A ), the set-point amplitude 
(A sp ), the tip shape, the cantilever force constant, the compliance of the sample to the tip, 
and the frequency of the oscillation [113]. Correlating regions of different phase shifts 



132 

with the corresponding polymeric phases requires knowledge of type of sample-tip 
interaction that is occurring during imaging. In fact other experiments are often needed 
to correlate the polymeric components with the areas of different phase shift. The types 
of interactions that take place are partially controlled by the feedback parameters used to 
image the surface. Three regions of tip-sample interaction have been identified based on 
the ratio of the set-point amplitude (A sp ) to the free oscillation amplitude (A ), r sp = 
(A sp /A ). The set-point amplitude is the amplitude that is maintained during scanning by 
adjustment of the vertical position of the tip. The free-oscillating amplitude is the 
amplitude of the AFM tip oscillation when it is not in contact with the surface. The 
regions are as follows: soft tapping (r sp > -0.8), moderate tapping (r sp = ~0.7 - 0.2), and 
hard tapping (r sp < ~0.2)[1 14]. Van der Waals and capillary forces govern the soft 
tapping regime tip-samples interactions, and negative phase shifts are observed. In the 
moderate and hard tapping regime repulsive forces govern tip-sample interactions, and 
positive phase shifts are observed. It should be noted that in all regimes the cantilever 
experiences both repulsive and attractive forces at various points in its oscillation cycle. 
However, the forces discussed above are the dominant forces in that regime. The 
magnitude of the free oscillation amplitude also has an effect on the tip-sample 
interactions [1 14, 1 15]. At small amplitudes, approximately 15 nm, attractive forces are 
dominant at all r sp values. Larger amplitudes, approximately 45nm or greater, repulsive 
interactions dominate and positive phase shifts occur. Bar et al. have examined poly 
(ethene-co-styrene)/poly (2,6-dimethyl-l,4-phenylene oxide) blends by imaging samples 
in the soft, moderate and hard tapping regimes[l 14]. Contrast inversion occurred twice 
while the different regimes of tip-sample interaction were examined. In terms of the 



133 

images presented in this work, softer phases can be made to appear dark or light 
depending on the parameters selected. In the work by Bar et al. hard phases appear 
brighter, a greater positive phase shift, in the moderate tapping regime. Other work has 
also shown similar results [115-117]. In the moderate regime, the tip-sample interaction 
is dominated by sample stiffness; hence, the harder phases appear brighter. In the hard 
tapping regime soft phases appear brighter and hard phases appear darker. In the hard 
tapping regime the AFM tip spends a larger portion of its time in contract with the sample 
surface. Area effects result when the tip spends a significant portion of its oscillation in 
contact with the sample. A softer sample allows more material to come in contact with 
the tip of the sample. The larger sample area results in a larger effective stiffness. The 
higher stiffness results in increased repulsion forces thus a greater phase shift than the 
stiffer materials. Similar results have been demonstrated with poly (ethylene) blends 
composed to low and high density regions, and poly (diethyl siloxane) on silicon 
substrates [115]. 

Even with the knowledge of the different regimes of tip-sample interaction 
caution must be maintained when making assignments to regions of different contrast. 
Each polymer system must be treated individually. The response of known polymer 
systems in the different regimes of tapping should only be used as a guide for 
interpretation of images. As stated previously other experiments are often needed to 
correlate the polymeric components with the areas of different phase shifts. Two-studies 
on block copolymers illustrate these statements [1 13, 118]. The PMMA phase in a 
PMMA-b-PBD-b-PMMA polymer was identified as the component with less phase shift. 
This assignment was based on two correlations. First, a correlation was drawn between 



134 

the mass percent of phases and the image analysis. Second, a correlation was drawn 
between phase images and the corresponding height images. It was proposed that the 
softer domains, which correlate with regions of larger phase shift, appeared higher in the 
topography image. The increase in height is a result of softer phase polymer relaxing by 
protruding out of the surface of the material. The exact parameters of the tapping were 
not provided so the tapping regime that the image was recording in could not be 
ascertained. An assignment based on correlations between height and phase images 
alone is suspect because height artifacts are possible as a result of tip-sample interactions 
with materials of different stiffness [1 14]. In the moderate tapping regime stiff regions of 
a sample appear higher than the soft regions because they deform less during the tip- 
sample interaction. The feedback mechanism of the TMAFM adjusts the tip to sample 
distance so that constant amplitude is maintained. A hard material deforms less and 
results in the tip being adjusted higher relative to a soft sample to maintain the set-point 
amplitude. Conversely, a softer material allows the tip to indent more; therefore, the tip 
must be held at a lower height to maintain the same set-point amplitude. In a study of 
butadiene/styrene-co-butadiene rubber blends [1 13], the butadiene phase was identified 
as the phase with the greater phase shift. Although the styrene-co-butadiene phase is 
stiffer, area effects resulted in a greater phase shift for the butadiene phase. Frequency- 
sweep and force-probe measurements were performed to understand the factors affecting 
the images. Force-probe experiments indicated that at any given r sp value the indention 
depth of the tip is greater in the butadiene phase. The more compliant sample allows a 
greater amount of materials to come in contact with the tip. This results in a greater 
phase shift. Correlation of indention depth to phase shift showed that at a constant 



135 

indention depth the styrene-co-butadiene phase had the greater phase shift due to its 
higher stiffness. Similar results were reported in poly (dimethylsiloxane) samples 
formulated with different cross-links densities [119, 120]. Despite the varying moduli 
there was not a significant difference in the phase shift for PDMS sample with different 
cross-link densities. This similarity was attributed to area effects. 

Initially fracture surfaces were imaged to avoid surface and mold effects and to 
expose the bulk morphology. The images of the fracture surfaces of 60/40 Bis- 
MEPP/TEGDMA (Figures 4.2 and 4.3) indicate that the morphology is on the range of 25 
to 150 nm. The light and dark areas represent the low and high cross-link density phases. 
Assignment of the regions of different contact will be addressed later in this work. The 
tortuous surface of the fractured samples prevented extensive imaging. Microtomed 
samples provided a less tortuous surface while still exposing the bulk morphology of the 
samples. Images taken of the microtomed surfaces show the same features as those 
samples prepared by fracture. The topographic and phase images of a microtomed 70/30 
Bis-MEPP/PEMA sample at a 500 nm scale (Figure 4.4) was recorded with a free 
oscillating amplitude (A,) of 3.0 V and set-point amplitude of approximately 1.0 V. 
These parameters correspond with what has been described previously as moderate 
tapping. The same sample at a 1 micron scale (Figure 4.5) was recorded with an A value 
of 3.0 V and a set-point values of 0.78 V. These settings fall within the low end of r sp 
values that have been defined as moderate tapping. The terms Ao and A sp have been 
discussed previously in terms of amplitudes. A and A sp are given in units of voltage by 
the AFM software. The voltages reported by the instrument are related to the 



136 



500 



250 




500 



250 



100.0 deg 



50.0 deg 



'0.0 deg 



NanoScope Tapping AFM 
Scan size 500.0 nM 

Setpoint 1.477 U 

Scan rate 3.590 Hz 

NuMber of saMples 512 



100.0 nn 



50.0 nn 



'0.0 nM 



NanoScope Tapping AFM 
Scan size 500.0 nM 

Setpoint 1.477 U 

Scan rate 3.590 Hz 

NuMber of saMples 512 



nM 



Figure 4.2: Phase and topographic AFM images of a fracture surface of a 60/40 Bis- 
MEPP/TEGDMA polymer at 500 nm scale 









137 




1.00 



0.75 



0.50 



100.0 dec, 



50.0 deg 



■0.0 deg 



1-0,25 NanoScope 
Scan size 



Tapping AFM 

1.000 mm 
Setpoint 1.477 U 

Scan rate 3.590 Hz 

NuMber of sanples 512 




0.25 



0.50 



0.75 



1.00 



0.75 



0.50 



■150.0 nn 



75.0 hm 



■0.0 nM 






0.25 NanoScope Tapping AFM 

Scan size 1.000 dm 

Setpoint 1.477 U 

Scan rate 3.590 Hz 

NuMber oT sanples 512 




1.00 

MM 



Figure 4.3: Phase and topographic AFM images of a fracture surface of a 60/40 Bis- 
MEPP/TEGDMA polymer at a 1 urn scale 



138 

amplitude of the tip oscillation; however, the actual amplitude of oscillation depends on 
the spring constant of the tip. A separate experiment is necessary to report the magnitude 
of the oscillation. The different regions of phase shift (Figure 4.4) represent the high and 
low cross-link density phases of the polymer. Examination of phase and topographic 
images shows that at a large scale, the height variation from the lower-left hand corner to 
the upper-right hand corner, the phase image is independent of the topographic features. 
At a smaller scale there is a correlation between the bright features and the local 
maximums in the topographic image. There is no sharp change in contrast between the 
light and dark regions and there also appears to be some regions with an intermediate 
contrast. The lack of sharp contrast change is most likely a result of domains that are 
partially located below the sample surface. Similarly, the regions of intermediate contrast 
are a result of phases that are located below the exposed surface. The 1000 nm image 
(Figure 4.5) has the same features as the 500 nm image (Figure 4.4) except the regions 
scale down with the size of the image. The size of the regions seen in the phase images 
correlates with the size of polymeric particles that have been isolated early in the cure 
cycle of ethylene glycol dimethacrylate[l 1]. The isolated particles had a biomodal size 
distribution with mean sizes in the range of 19 to 30 nm and 83 to 134 nm. It was 
proposed that larger particle sizes might be a result of the agglomeration of the primary 
particles. Strict comparison between the ethylene glycol dimethacrylate system and the 
ones examined this work is not possible because of the differences in the monomers 
structure and the initiator systems; however, the correlation between sizes of the isolated 
particles and the domains sizes seen in the phase images is encouraging. It should also be 
noted that it has been suggested that cross-linked polymers formed by chain 



139 




500 



250 



,100.0 deg 



50.0 deg 



1 . deg 



NanoScope Tapping AFM 
Scan size 500.0 dm 

Set point 0.9942 U 

Scan rate 3.052 Hz 

NuMber of saMples 512 




500 



250 



hm 



100.0 nM 



50.0 nM 



'0.0 nM 



NanoScope 
Scan size 
Setpoint 
Scan rate 



Tapping AFM 

500.0 nM 
0.9942 U 
3.052 Hz 



NuMber of saMples 



512 



Figure 4.4: Phase and topographic AFM images of a microtomed surface of a 70/30 Bis- 
MEPP/PEMA polymer at 500 nm scale 



140 




1.00 



0.75 



.50 



100.0 deg 



50.0 deg 



■0.0 deg 



0.25 NanoScope Tapping AFM 

Scan size 1.000 mm 

Setpoint 0.7692 U 

Scan rate 3.052 Hz 

NuMber of sanples 512 



MM 




0.25 



0.50 



0.75 



1.00 



0.75 



0.50 



100.0 nM 



50.0 nM 



■0.0 nM 



0.25 NanoScope Tapping AFM 

Scan size 1 .000 mm 

Setpoint 0.7692 U 

Scan rate 3.052 Hz 

NuMber of saMples 512 





1.00 

MM 



Figure 4.5: Phase and topographic AFM images of a microtomed surface of a 70/30 Bis- 
MEPP/PEMA polymer at 1 urn scale 



141 

polymerization have areas of low and high cross-link density with sizes on the order of 
10 to 50 nm [1 10]. The variation in phase size is notable especially in the 1 micron 
images. This variation is expected due to the complex mechanism of formation of these 
polymers. As discussed early in this work heterogeneity in dimethacrylate polymers has 
it roots early in polymerization with the formation of microgel particles. Briefly, 
preferential polymerization around these particles and their eventual agglomeration 
results in the region of low and high cross-link density, heterogeneity. The variation in 
the domain sizes of the structures is a result of how the polymer particles agglomerate 
during polymerization. 

As discussed previously, assigning regions of different contrast to specific 
morphological structures is a complex task and additional knowledge of the system in 
question is required to assign regions of different phase contrast to specific morphologies. 
To help make the assignments a microtomed sample was immersed in acetone for 36 
hours, then removed and dried, and a post cured sample was also prepared. The phase 
and topographic images of the sample treated with acetone are contained in Figure 4.6. A 
3-D perspective image of the phase image is provided in Figure 4.7 to elucidate 
morphology change induced by exposing the sample to acetone. Based on the images of 
the light-cured samples an assignment of specific morphological regions to area with 
different phase shift is not possible. The first factor considered was the relative amount 
of the different phases present. Based on the degree of conversion (-66%) of the 70/30 
Bis-MEPP/PEMA, it might be expected that there would be a ratio of approximately 7 to 
3 high cross-link density phase to the low cross-link density phase. However, it is not 
possible to make this assumption because the relative concentration of double bonds 



142 

within the different phases is not known. Chiu et al. have shown that the isolated 
microgels of ethylene glycol dimethacrylate, which is considered to be a region of high 
cross-link density, were found to have pedant double bonds [11]. One correlation that 
can be drawn from the images in Figures 4.2-4.5 is the relationship between local maxim 
in the height image and regions of bright contrast or large phase shifts. It has been 
proposed that a softer phase has the ability to relax by protruding out of the surface of a 
sample due to its higher mobility [118]. Based on this analysis it is temping to make the 
correlation between the regions of high phase shift and the low cross-link density phase. 
Two other factors must be considered that impede the assignment of the regions of high 
phase contrast to the low cross-link density phase. One, it has been shown that harder 
phases appear artificially high in topographic images due to reasons discussed earlier 
[1 14]. Two, examination of the images of samples exposed to acetone (Figures 4.6 and 
4.7) shows that the regions of high phase shift correlate with higher topographic features. 
The exposure of the samples to acetone could result in the removal of the low cross-link 
density material or preferential swelling of the low cross-link density phase. Comparison 
of a post cured sample to a standard light-cured sample give another point of comparison 
to make the assignment. The phase and topographic images of a standard 70/30 Bis- 
MEPP/PEMA sample and the post cured 70/30 Bis-MEPP/PEMA (Figures 4.8 and 4.9) 
were both taken with the same tip and feedback parameters to minimize artifacts caused 
by differences in tip spring constants and imaging in different tapping regimes. The post 
cure should reduce the amount of volume of the low cross-linked density phase. The size 
of the region that represents that low cross-link density phase will be reduced in the 
images. In the images of the post cured sample (Figure 4.9), the lighter region (higher 



143 




500 



250 



100.0 deg 



50.0 deg 



■o.O deg 



NanoScope Tapping AFM 
Scan size 500.0 nn 

Setpoint 0.9121 U 

Scan rate 2.636 Hz 

NuMber of saxples 512 



nM 




500 



100.0 nM 



250 



50.0 nn 



'0.0 nM 



NanoScope 
Scan size 
Setpoint 
Scan rate 



Tapping AFM 

500.0 nM 
0.9121 U 
2.636 Hz 



NuMber of saMples 



512 



Figure 4.6: Phase and topographic AFM images of a microtomed surface of a 60/40 Bis- 
MEPP/TEGDMA polymer immersed in acetone and then dried 



144 



NanoScope 
Scan size 

Setpoint 
Scan rate 



Tapping AFH 

500.0 rin 
0.9121 U 
2.636 Hz 



Number of saxples 



512 




100.000 nM/div 
100.000 deg/diu 



Figure 4.7: The 3-D perspective of an AFM phase image of a microtomed surface of a 
60/40 Bis-MEPP/TEGDMA polymer immersed in acetone and then dried 



145 

phase shift) increases in size and the darker region (lower phase shift) decreases. This 
indicates that the brighter regions (higher phase contrast) represent the higher cross-link 
density phase. The growth of the bright regions (greater shift phase) upon post curing 
coupled with the apparent etching of dark phase (low phase contrast) in the samples 
exposed to acetone allows the brighter areas to be assigned to the high cross-link density 
phase. The assignment of the brighter regions to the high cross-link density phase 
correlates well with other studies, which showed that stiffer regions had a larger phase 
shift in the moderate tapping regime[l 14, 116]. 

It should be noted that there is some difference in appearance between the 70/30 
images with a A sp of 0.4 V (Figure 4.8) and the one imaged with a A sp of- 1.0 V (Figure 
4.4). The phase image taken at the higher set-point (Figure 4.8) appears to have more 
fine features than the image taken at the lower set-point (Figure 4.4). Some of the 
difference may be due to local variations in microstructure; however, the image in Figure 
4.8 was taken in the hard tapping regime. The stronger sample tip interaction results in 
greater exposure of phases that are located beneath the surface of the sample. Phase 
inversion, the reversal in contrast between the different phases, has been reported when 
samples are imaged in different tapping regimes [1 14], [1 15]; however, phase inversion 
does not occur in all systems [113]. There were no signs of phase inversion observed in 
the images obtained in this work. 
Xenon- 1 29 NMR 

The narrow peak of the xenon NMR spectra of the Bis-MEPP(XDT) monomer 
(Figure 4.10) is an indication that the xenon samples all the chemical environments in the 
monomer at a high frequency. The fast exchange results in a signal that is an average of 



146 




500 



250 






100.0 deg 



50.0 deg 



0.0 deg 



NanoScope Tapping AFM 
Scan size 500.0 rm 

Setpoint 0.4000 U 

Scan rate 2.521 Hz 

NuMker of sanples 512 




500 



250 



50.0 nM 



25.0 nn 



■o.O nn 



NanoScope Tapping AFM 
Scan size 500.0 nM 

Setpoint 0.4000 U 

Scan rate 2.521 Hz 

Number oT saMples 512 



Figure 4.8: Phase and topographic AFM images of a microtomed surface of a 70/30 Bis- 
MEPP/PEMA polymer imaged with a set-point value of 0.40 volts at 500 nm scale 






147 




500 



250 



100.0 deg 



50.0 deg 



■0.0 deg 



NanoScope Tapping AFM 
Scan size 500.0 dm 

Setpoint 0.4000 U 

Scan rate 2.521 Hz 

NuMber of sanples 512 




500 



250 



nM 



100.0 nn 



50.0 nn 



'0.0 nM 



NanoScope Tapping AFM 
Scan size 500.0 nM 

Setpoint 0.4000 U 

Scan rate 2.521 Hz 

NuMber of saMples 512 



Figure 4.9: Phase and topographic AFM images of a microtomed surface of a 70/30 Bis- 
MEPP/PEMA polymer post cured for 2 hours at 140°C immediately after light-curing 
imaged with a set-point value of 0.40 volts at 500 nm scale 



148 

all the environments. It should be noted that a comparison of the relative magnitudes of 
the peaks is not possible for Figures 4.10-4.12. One can easily discern that the peak 
shifts further downfield as the degree of conversion is increased. The peak width 
increases and the intensity decreases with longer cure times. Similar behavior was 
reported when xenon NMR was used to follow the cross-linking reaction of poly 
(urethane) [81]. As the Bis-MEPP(XDT) system cures and becomes more dense and less 
uniform the exchange of xenon between the different environments slows down. The 
greater distribution of environments and the smaller 



22 


3,530,106 


274.5 


941,252 


545.8 


298,921 



Table 4.5: Xenon NMR spectrum parameters for Bis-MEPP(XDT) at various stages of 

cure 

Sample Chemical Shift (ppm) Line Width Intensity 

(Hz}_ (AU) 

Monomer 190.0 

30 195.3 

5 minutes 198.9 



amount of signal averaging result in a broader peak. Similarly, the intensity of the peaks 
decreases from 3,530,106 for the monomer to 298,921 for the sample cured for 5 
minutes. As the Bis-MEPP cures the high cross-link density phases compose more 
volume fraction of the sample and the xenon is expelled from the sample. The smaller 
amount of xenon present in the sample results in lower signal intensity. 

The sample cured for 5 minutes has attained the maximum degree of conversion 
possible by light curing at room temperature. A Bis-MEPP sample cured using the 
camphorquinone/di-methyl-p-toluidine initiator described elsewhere in this work attained 
roughly the same degree of conversion. Upon removal from the NMR tube the sample 



149 



HP 






on 






b~Fi 






IT; 






Ol 






5ji 






oTi 






oT; 






ST; 






b"T~! 


...... 




D.0 | 


' ' I ' ' ' ' I ' ' ' ' I ' ' ' ■ I ' ■ ' ' 1 ■ ' ' ' 1 ....,...., 



200 150 100 50 

Figure 4.10: Xenon- 129 spectra of Bis-MEPP(XDT) monomer at 25°C 



150 

appeared to be glassy. The xenon spectra of glassy polymers are generally weak and very 
broad. The line width of polycarbonate was reported at 500 Hz in a 4.7 tesla magnetic 
field strength [77]. The line width of PMMA was reported at 2000 Hz at 4.7 tesla 
magnetic field strength. The full width half maximum line width of the 5-minute cure 
sample was 545.76 at 9.4 tesla magnetic field. Assuming a rapid exchange limit is 
reached the change in line- width is proportional to the square root of the magnetic field 
strength [77]. This approximation estimates the line width of the 5-minute cure sample at 
385 Hz in a 4.7 tesla magnetic field. The smaller line width, line shape, and intensity of 
the spectrum indicate a portion of the material in this sample is not in a fully vitrified 
state and has a pore structure that accommodates more xenon than a fully glassy 
structure. The existence of a bimodal distribution of pore structures is corroborated by 
the bimodal phase distribution seen in the AFM images presented earlier in this chapter. 
The lack of two clearly resolved peaks in the spectra of the samples cured for 30 
seconds and 5 minutes is not necessarily an indication of a material consisting of a single 
phase. As mentioned earlier the solubility of xenon in the glassy highly cross-linked 
phase is small compared to that of the monomer and the low cross-link density phase. 
The high density results in the low xenon solubility and a broad weak signal. The 
contribution of the highly cross-linked phase to the signal is minor; therefore, the signal 
of the highly cross-link phase cannot be easily resolved from that of the low cross-linked 
phase. Depending on the rate of the diffusion of xenon in the various phases and the size 
of the phases, the single peak might be the result of signal averaging from the different 
phases. Similar behavior has been reported in PEO/PMMA blends [87]. PMMA, a 
glassy polymer, was referred to as a "hidden exchange partner" because its contribution 



151 


ol 






o~i 






on 






crj 






01 






o~i 






on 






oTj 






on 




' 


O"; 


M 


Wm^^ 


0.D ; 


-^f^" 


^l^ll(^IW^ftf l f^^^ , n^f^W^T^^ n r HT™ ™Wm 




■ ■ ■ ■ 1 1 ,....,...., ......... 

200 150 100 50 


Figure 4.1 1: Xenon- 129 spectra at 25°C of Bis-MEPP(XDT) cured for 30 seconds at 


25°C 















152 

to the signal in the blends was difficult to resolve. As mentioned earlier, PMMA has a 
broad weak signal. 

If two assumptions are made, an estimation of the volume fraction of the phases 
that are present can be made based on the difference in intensity of the monomer 
spectrum and the spectrum of the sample cured for 5 minutes. The first assumption is 
that the differences in pressures between the xenon samples in negligible and does not 
affect the intensity of the signal. The second assumption is that the two phases formed 
are a highly cross-linked phase, in which gas solubility is negligible compared to the 
monomer phase, and a monomer like phase. The volume fraction of monomer like phase, 
calculated using these assumptions, is approximately 8.5 volume percent. A more 
accurate description of the monomer phase is one of low cross-link density with higher 
mobility than the highly cross-linked phase. The solubility of xenon in this phase is 
lower than that of the monomer. This indicates that the volume fraction of the low cross- 
link density phase is greater than 8.5 percent. 

As might be expected the self-diffusion coefficients of xenon decrease with 
increasing conversions (Table 4.6). The diffusion coefficients in the Bis-MEPP(XDT) 
samples were not measurable in the 30-second and 5-minute cure samples for reasons 
that will be discussed later. 15 urn is the maximum domain size measurable as indicated 
by the results in the monomer system. In the system cured for 20 seconds that value has 
been reduced to approximately 12 um. This is due to the shorter mean free path in the 
monomer like environment due to the formation of high cross-link density phases and the 
increased density of the systems. The samples cured 10, 15, and 20 seconds all had gel- 
like physical properties upon removal from NMR tubes. The sample cured for 30 



153 







-0.1 -i 

Figure 4 



200 150 100 SO 

12: Xenon- 129 spectra at 25°C of Bis-MEPP(XDT) cured for 5 minutes 



154 

seconds can be described as leathery. The physical properties of the sample cured for 5 
minutes appeared to be glassy. The critical conversion for gelation in tetrafunctional 
systems predicted with the following equation based on percolation theory is 
approximately 33 percent [121]: 



c (/-l) 






Where: P c = critical conversion to gelation 

/ = average functionality of the system 
The sample cured for 30 seconds had a degree of conversion of approximately 26 
percent. The ability to make pulsed field gradient measurements ceased and the physical 
properties of the sample changed from gel-like to leathery in the 30-second cure samples. 
This suggests the gel formation is related to these changes in properties. The similarity of 
the values for the systems cured for 10, 15, and 20 seconds may be due to some depth of 
cure issues. Initially, there is probably a higher concentration of radicals in the monomer 
near the external region of the sample than in the center. After 30 seconds of curing the 
radical concentration may be more uniform. This may account for the jump in 
conversion between the sample cured for 20 seconds and the sample cured for 30 
seconds. 

Diffusion coefficients were not measurable in the systems cured for 30 seconds 
and 5 minutes. A signal was not observed using the original parameters of the stimulated 
echo sequence. A modified sequence as described in the materials and methods section 



155 



Table 4.6: Degree of conversion values, diffusion coefficients, and domain sizes of Bis- 
MEPP(XDT) at various stages of cure 



Cure Time 


Degree of 


Diffusion 


Domain Size 


(Seconds) 


Conversion 


Coefficient 


Measured 




(%) 


(M 2 /S) 


(Mm) 


Monomer 





7.0 ± 0.53 xlO" 11 


15 ±0.6 


10 


3.3 ±1.7 


5.4 ±0.18x10"" 


13 ±0.3 


15 


5.5 ±2.6 


5.2 ± 0.16 xlO" 11 


13 ±0.2 


20 


7.6 ±2.4 


5.1 ±0.21 xlO" 11 


12 ±0.3 


30 


26 ±3.0 


N/A 


N/A 


5 min 


54.1 ±3.4 


N/A 


N/A 



was used. The time between the pulses and the length of gradient pulses was decreased 
to prevent loss of signal due to T2 or spin-spin relaxation. A signal was observed, but no 
attenuation occurred within the noise of the spectrum. Similar difficulty was encountered 
with the sample cured for 5 minutes. The root of the problem is signal attenuation due to 
rapid T2 relaxation. Assuming that only homogeneous signal broadening is occurring, 
the T2 relaxation time of a system can be estimated using the following equation [122]: 



&&,/„ =4- 



1/2 T 2 



Where 8©i/2 = the spectrum width and half-maximum in hertz. 
The T2 relaxations times of the systems cured for 30 seconds and 5 minutes are 7.3 and 
3.7 milliseconds, respectively. For comparison, the T2 of the monomer systems is 90 
milliseconds. In the second pulse sequence described for the 30-second cure sample 
(Table 4.5) the time elapsed between the first nil pulse and the second is nearly 3 
milliseconds. During that period a large portion of the signal has been lost for the 30- 
second cure sample and nearly all of the signal has been lost for the 5-minute cure 



156 

sample. As mention early, there was no resolvable signal attenuation for 30-second cure 
sample. This is due to the short duration of the gradient pulses (0.5 ms). In other studies 
where gradient pulses of this duration were used, gradient strengths of nearly an order of 
magnitude greater were required to attain measurable signal attenuation [86]. 
Unfortunately, our gradient strength is already at its maximum; however, this does 
suggest that these measurements can be pushed further into the cure cycle with the 
appropriate equipment. 

Conclusions and Future Work 

Tapping Mode™AFM measurements using phase imaging demonstrated that 
there are variations in the cross-link density of dimethacrylate-type polymers on the order 
of 25-150 nm. The spectrum of Bis-MEPP in its fully cured state corroborates the AFM 
results. The spectrum indicates that there are lower cross-link density regions present 
with a pore structure that accommodates xenon despite having a macroscopic glassy 
appearance. The presence of a structure with a bimodal distribution of cross-link 
densities is also indicated by the DMS data presented in Chapter 3. These results all 
show that dimethacrylate polymers are heterogeneous materials composed of regions of 
low and high cross-link density rather than a continuum-type structure. 

Further refinement of the AFM techniques may allow relationships between the 
sizes of the different regions and mechanical properties to be established. The local 
variation in domain sizes in the images makes it difficult to determine their average size 
by casual observation. Image analysis software would be useful to determine the average 
size of the domains as well as their volume fractions and to establish relationships 
between mechanical properties and domain sizes. 



157 

Pulsed field gradient xenon NMR was used to measure the change in domain 
sizes with cure in dimethacrylate systems. Unfortunately, measurements could only be 
made at conversions less than 26 percent. Equipment in which stronger field gradients 
can be applied would allow the measurement to be pushed further into the cure cycle. In 
polymer blends where peaks of the different phases are resolved, changes in the exchange 
rate can observed by 2D experiments as well as changes in the ID spectra such as 
coalescence of the peaks. If the xenon self-diffusion coefficient of the system is known, 
then the exchange rate can be used to calculate domain sizes. This type of analysis is not 
possible when one of the phases is acting as a hidden exchange partner. This means that 
the alterations in the signal due to different exchange rates are not visible. Schantz and 
Veeman have measured domain sized on the order of 50 nm in systems where one phase 
is acting as a hidden exchange partner by modeling the one-dimensional spectra of the 
system using the exchange rate as a variable. A similar analysis might be applied to 
dimethacrylate systems to estimate domain sizes to corroborate those observed by AFM. 



CHAPTER 5 

EVALUATION OF THE EFFICIENCY OF THE INCORPORATION OF NADIC 

METHYL ANHYDRIDE, NORBORNENE BASED COMPOUNDS AND MALEIC 

ANHYDRIDE INTO METHACRYLATE-BASED DENTAL RESINS 



Relevant Background 
Monomers that introduce cyclic groups into the methacrylate backbone of dental 
polymers, such as maleic anhydride, methylene lactone, and norbornene-type monomers, 
have been studied with mixed results. This chapter will investigate the mechanistic and 
structural behavior of nadic methyl anhydride and maleic anhydride polymerized with 2- 
phenyl ethyl methacrylate. The properties of methylene lactone monomers, described as 
cyclic analogs of methyl methacrylate, lead to their study as coreactants with dental 
monomers [123]. Poly (a-methylene-y-butyrolactone) (MBL) (Figure 5.1) has a glass 
transition temperature of 195 °C as compared to MMA, which is approximately 105 °C. 
MBL is also more polar than MMA and is considered more reactive as determined by 
copolymerization studies with MMA. The addition of up to 20 weight percent MBL to 
Bis-GMA resins increased the diametral tensile strength of the polymer, which was 
conditioned for 24 hours at 37°C in water. The increase in properties is attributed to the 
lower viscosity of MBL as well as the incorporation of the rigid lactone rings into the 
methacrylate backbone. FTIR studies revealed that the addition of 30 weight percent 
MBL to Bis-GMA resins increased the conversion of vinyl bonds to 71% compared to 
32% for pure Bis-GMA. The addition of 10 weight percent MBL to Bis- 
GMA/TEGDMA resins increased the conversion of double bonds from 57 to 71 



158 



159 



o 




h 2 <t o 
i i 

C CH 2 

// 

H 2 C 

Figure 5.1: The structure of a-methylene-y-butyrolactone 




Figure 5.2: The structure of nadic methyl anhydride (NMA)(methyl-5-norbornene-2,3 
dicarboxylic acid anhydride) 



°- o ,P 

=CH 




Figure 5.3: The structure of maleic anhydride (MA) 



160 

percent compared to Bis-GMA/TEGDMA systems. A 10 percent increase in strength 
values also occurred; however, the increase was not statistically significant at a = 0.05. 
The increase in conversion could be a result of the higher reactivity of the MBL. The 
lower initial viscosity and average functionality of MBL containing mixtures are factors 
as well. Improved mechanical properties have been reported in Bis-MEPP/TEGDMA 
systems in which 10 wt.% nadic methyl anhydride (NMA) was added. Despite its rigid 
ring structure, the addition of 20 wt. % nadic methyl anhydride (NMA) to Bis- 
MEPP/TEGDMA systems (60/40 weight percent) lowered the glass transition 
temperature by 42 °C) (Figure 5.2)[99] compared to the Bis-MEPP/TEGDMA (60/40 
weight percent) systems. Similarly, MLB may lower the glass transition temperature of 
the systems in which it is incorporated by lowering the cross-link density; although, its 
incorporation probably stiffens the methacrylate backbone of the polymer. Maleic 
anhydride (MA) (Figure 5.3) incorporated into urethane dimethacrylate/TEGDMA 
composites at 30 mol.% lowered the flexure strength and modulus by 49 MPa and 2.0 
Gpa, respectively. [64]. The incorporation of maleic anhydride into 2-phenylethyl 
methacrylate polymers resulted in higher glass transition temperatures [124]. One might 
expect similar results from the MBL and NMA systems as from the MA systems. The 
incorporation of MA into these 2-phenylethyl methacrylates was shown to be inefficient 
by FTIR and DSC, indicating a low reactivity of MA with methacrylates. NMA is also 
assumed to be less reactive due to steric considerations. The extractable sol fraction of 
Bis-MEPP/TEGDMA/NMA polymers increased with increasing NMA concentration, 
which indicates a lower reactivity of NMA with MEPP/TEGDMA. This could be a result 
of the lower functionality of NMA. 






161 

This work will attempt to understand the findings in nadic methyl anhydride 
modified methacrylate systems by determining the efficiency of its incorporation along 
with other norbornene-based monomers and maleic anhydride. The model compounds 
were based on the monomers 2-phenyethyl methacrylate (PMA), nadic methyl anhydride, 
and maleic anhydride. PMA is model compound for aromatic dimethacrylates such as 
ethoxylated bis-phenol dimethacyrate and Bis-GMA. PMA forms linear polymers that 
are more easily characterized than cross-linked systems. The compounds were 
characterized using FTIR, GPC, DSC and 'H NMR. The effect of anhydride feed 
concentration on their incorporation into the PMA based polymers was determined. The 
incorporation of maleic anhydride and nadic methyl anhydride through a reaction in the 
vinyl bond was also confirmed. 

Materials and Methods 
Model compounds were made by polymerizing various norbornenes with 2- 
phenylethyl methacrylate (PMA) (Figure 5.4). The norbornene compounds that were 
examined were methyl-5-norbornene-2,3 dicarboxylic acid anhydride or nadic methyl 
anhydride (NMA) (Figure 5.2), 5-norbornene-2,3 dicarboxylic acid anhydride (Figure 
5.4), 5-norbornene-2-carboxaldehyde (Figure 5.5), and 5-norbornene-2-butane (Figure 
5.6). Norbornene monomers were added at 10, 20, 30, and 40 mol.% concentrations to 
PMA. Model compounds that contained nadic methyl anhydride and maleic anhydride 
(MA) were also formulated (Table 5.1). PMA was vacuum distilled prior to use at 72 °C 
under a vacuum (< ImmHg). Maleic anhydride was recrystallized from benzene. The 
compounds were polymerized in bulk using 0.4 weight percent AIBN as an initiator. The 
monomer/initiator mixtures were purged with dry nitrogen, sealed, and then polymerized 



162 

at 75°C for 4 hours. The polymer was then dissolved in chloroform and precipitated in 
methanol. Samples were filtered and then dried under vacuum (28 in Hg) at 40°C. 



Table 5.1: 2-phenylethyl methacrylate/nadic methyl anhydride/maleic anhydride 

monomer compositions in mol.% 

Sample PMA NMA MA 
PNM1 70 22.5 7.5 

PNM2 70 15 15 

PNM3 70 7.5 22.5 



The molar masses were determined by GPC using a Waters HPLC system 
composed of a Waters 600 Fluid Delivery System, a Waters 717 Auto sampler, a Waters 
410 Differential Refractometer, and a Waters 996 Photodiode Array Detector. Four 
Phenomenex cross-linked polystyrene columns with pore sizes of 10 5 ,10 4 , 500, and 100 
A were used in series at a flow rate of 0.4 ml/min. The injection volume was 50 ul of 
each sample, which were approximately 0.25 weight percent in HPLC grade THF. 
Anionically polymerized polystyrene standards obtained from Polymer Laboratories were 
used for calibration. The samples and standards were also randomized. 

A Seiko DSC 220C Differential Scanning Calorimeter interfaced with a Seiko 
SDM/5600h Rheostation was used to determine the glass transition temperatures of the 
model compounds. The temperature of the sample and an inert alumina reference are 
monitored with thermocouples. Samples were heated at a rate of 10°C/min. under a 
nitrogen purge of 100 ml/min. Samples were scanned from -50°C to 150°C three times. 
A baseline thermogram was run with an empty pan and subtracted from the sample 
thermograms. Three runs were made on all of the compositions tested. 



163 

The polymers prepared with nadic methyl and maleic anhydrides were examined 
with Fourier Transform InfraRed spectroscopy (FTIR) to determine the concentration of 
anhydride in the various polymers. Calibration standards for the anhydride groups were 
made from the infrared spectra of the monomer mixtures prior to the addition of catalysts. 
Peaks at 1781 cm" 1 and 1853 cm" 1 due to the asymmetric and symmetric carbonyl 
stretches identify the anhydride functionality. The aromatic double bonds at 1608 cm" 
were used to identify the 2-phenylethyl methacrylate. Peak fitting software was used to 
deconvolute and measure the areas of these peaks (PeakFit v4 by SPSS Inc.). 

The specific site of NMA incorporation into the polymer was investigated with H 
Nuclear Magnetic Resonance Spectroscopy (NMR) of the model compounds to 
determine if the norbornenes incorporated via reaction through the double bond. The 
NMR spectroscopy was performed on a Gemini 300 NMR system composed of a 7 Tesla 
Oxford magnet and 5mm Varian broadband probe. The spectra were referenced to 
tetramethylsilane (TMS). The model compounds and base monomers were dissolved in 
deuterated-dimethylformamide deuterated at 2.5 weight percent. The solutions were 
made in a dry environment (less then 20% relative humidity) to avoid water uptake in the 
DMF. 

Results and Discussion 
A full study was only performed on the model compounds based on 2-phenylethyl 
methacrylate/nadic methyl anhydride and 2-phenylethyl methacrylate/nadic methyl 
anhydride/maleic anhydride compounds. Solubility issues were encountered with the 
other compounds preventing further analysis. The aliphatic norbornene compounds were 
soluble in the 2-phenylethyl methacrylate monomer; however, as the polymerization 






164 




o 
■0-C-C-CH3 

CH 9 



Figure 5.4: The structure of 2-phenylethyl methacrylate 







Figure 5.5: The structure of 5-norbornene-2,3 dicarboxylic acid anhydride 



165 




Figure 5.6: The structure of 5-norbornene-2-carboxaldehyde 







R= (CH 2 ) 3 CH 3 or (CH 2 ) 5 CH 3 



Figure 5.7: The structures of 5-norbornene-2-butane and 5-norbornene-2-hexane 



166 

progressed, phase segregation occurred. 5-norbornene-2,3 dicarboxylic acid anhydride 
(Figure 5.4) and 5-norbornene-2-carboxaldehyde (Figure 5.5) were not soluble in 2- 
phenylethyl methacrylate monomer. It should also be noted that 2-phenoxy ethyl 
methacrylate, the original monomer planned for this study, was discarded due to 
difficulties obtaining compounds which were soluble. 

The FTIR and DSC data collected from 2-phenylethyl methacrylate/nadic methyl 
anhydride type polymers indicates a low incorporation efficiency of NMA with PMA, 
i.e., less than 2 weight percent. The amount of NMA in the 2-phenylethyl methacrylate 
polymers is not sensitive to the concentration of NMA in the monomer feed (Table 5.2). 
Examples of FTIR from various systems are shown in Figure 5.8. As might be expected 
from the FTIR results, the glass transition temperature of the polymers was also not 
affected by the feed concentration of the monomers. The glass transition temperature of 
poly (norbornene) is in the range of 350°C [125]. Incorporation of NMA into the 
polymer at an amount greater than 2 percent would increase the glass transition 
temperature of the polymers formed. The increase in the feed concentrations of NMA 
resulted in no detectable difference in glass transition temperatures. 

The small amount of NMA incorporated does not change the molar mass of the 
PMA polymer, which implies that it does not alter the polymerization kinetics of the 
PMA. It should be noted that there is large batch-to-batch variability in bulk 
polymerization processes. This variability is illustrated by the differences in molar mass 
between the original samples and their replicates. The high amount of scatter may 
obscure any trends in molar mass or polydispersity that the presence of NMA might 
induce. 



167 



Table 5.2: FTIR and DSC results from 2-phenylethyl methacrylate/nadic methyl 
anhydri de based model compounds 



Polymer 


Tg(°Q 


Anhydride content 
(mol.%) 


Poly (PMA) 


31 ±4 





Poly (PMA-co-10%NMA) 


32 ±2 


Less than 2 


Poly (PMA-co-20%NMA) 


29 ±1 


Less than 2 


Poly (PMA-co-30-%NMA) 


33 ±6 


Less than 2 


Poly (PMA-co-40%NMA) 


31 ±4 


Less than 2 



Table 5.3: Molar mass averages from GPC for 2-phenylethyl methacrylate-nadic methyl 
anhydride copolymers 



Polymer 


Mn 
(kg/mol) 


PDI 


Poly (PMA) 


42.6 


3.6 


Poly (PMA) repeat 


58.9 


4.8 


Poly (PMA-co-10%NMA) 


41.2 


4.9 


Poly (PMA-co-10%NMA) repeat 


29.5 


4.4 


Poly (PMA-co-20%NMA) 


47.8 


4.7 


Poly (PMA-co-20%NMA) repeat 


48.3 


4.1 


Poly (PMA-co-30-%NMA) 


55.1 


4.5 


Poly (PMA-co-30-%NMA) repeat 


64.1 


4.1 


Poly (PMA-co-40%NMA) 


48.0 


4.8 


Poly (PMA-co-40%NMA) repeat 


50.6 


4.2 



NMA, maleic anhydride (MA), and 2-phenylethyl methacrylate were blended 
with the idea that a synergistic effect between the maleic anhydride and the NMA would 
increase concentration of both anhydrides in the copolymer. Maleic 
anhydride/norbornene polymerizations have been reported that produced alternating 
copolymers with yields ranging from 60-76% [126]. In the same studies, it was 



168 




Figure 5.8: FTIR spectra of PMA-co-40%NMA, PMA, poly (PMA-co-40%NMA), and 
poly (PMA-co-7.5%NMA-co-22.5%MA) 






169 

determined that tert-butyl acrylate concentration in the copolymers, i.e., tert-butyl 
acrylate/MA/NB, increased linearly with its monomer feed concentration, up to a 
maximum mole fraction of 0.2. There was no reduction in either the number average 
molar mass or polydispersity. 

The concentration of anhydride functionalities increases with an increase in the 
feed concentration of MA in poly (PMA-co-NMA-co-MA), which indicates that MA is 
more reactive with PMA than NMA. The incorporation is confirmed by a concurrent 
increase in the glass transition temperature of the polymers. Previous studies on 2- 
phenylethyl methacrylate-co-maleic anhydride polymers showed that maleic anhydride 
incorporated at roughly 45% of its feed in the monomer system [124]. In the poly (PMA- 
co-22.5 mol.% NMA-co-7.5 mol.% MA) system there is 4.6 mol.% anhydride. The feed 
concentration of MA was 7.5 mol.%. Based on the results from PMA-co-MA systems, 
the amount expected would be approximately 3.4 percent. It is not possible to determine 
the ratios at which maleic anhydride or NMA are incorporated by FTIR only that the 
presence of NMA results in increased amounts of anhydride in the polymer compared to 
what would be expected due to MA by itself. In the poly (PMA-co-7.5mol.%NMA-co- 
22.5mol.%MA) system, the total amount of anhydride incorporation is 7.7 mol.%. The 
amount of MA incorporated estimated by comparison with poly (PMA-co-MA) systems 
should be approximately 10.1%. This indicates some inhibition of the incorporation of 
MA by the NMA. The exact nature of the incorporation of NMA and MA cannot be 
discerned. These results indicate that MA and NMA do not copolymerize efficiently with 
PMA at any concentration. 



170 

As mentioned previously polymers of norbornene, maleic anhydride (MA), t- 
butyl acrylate (TBA) reported yields as high as 76%[126]. The difference between the 
cited study and the present study could be due to the polymerization methods, solubility 
differences, reactivity differences, and ratios of feed monomers. It should also be noted 
that PMA is a methacrylate and TBA is an acrylate, which also might result in differences 
in reactivity. The norbornene, maleic anhydride, t-butyl acrylate copolymers were 
prepared by polymerization in a THF solution at 65°C for 24 hours. Both the 
PMA/MA/NMA and the Norbornene/MA/TBA systems used AIBN as an initiator. The 
PMA/MA/NMA were polymerized in bulk at 75°C for 4 hours. The long reacting time 
may have allowed additional reaction of the MA and norbornene. In the same work 
Houlihan et al. [126] explored other cycloelfinic-maleic anhydride copolymers. A 
norbornene-based monomer with a bulky side group did not polymerize with maleic 
anhydride. In the absence of an explanation for the lack of polymerization, it may be 
reasonable to attribute it to steric hindrance. The anhydride ring in NMA might also 
cause steric hindrance and reduce the reactivity of NMA with maleic anhydride (MA) 
and PMA. 

Table 5.5 contains the GPC data from the PMA/NMA/MA systems. At the low 
concentrations of anhydride incorporation seen in this study the average molar mass of 
the PMA polymers is not affected by the incorporation of the anhydrides enough to see a 
trend within the scatter of the measurement. 



171 



Table 5.4: FTIR data and DSC from 2-phenylethyl methacrylate/nadic methyl 
anhydride/maleic anhydride based model compounds 



Polymer 


Tg (°C) 


Anhydride 
Content 
(mol.%) 




Poly (PMA-co-7.5%NMA-co- 


37 ±4 


7.7 ±1.1 




22.5%MA) 








Poly (PMA-co- 1 5%NMA-co- 1 5%MA) 


35 ±7 


5.110.9 




Poly (PMA-co-22.5%NMA-co- 


31+3 


4.6 ± 0.6 




7.5%MA) 









Proton NMR was performed to confirm that the anhydrides are incorporating 
through the polymerization of the vinyl functionality. It should be noted that in all 
spectra the peaks located at 8.03, 2.92, and 2.75 ppm are due to the DMF solvent. 



Mn 


PDI 


(Kg/mol) 




38.7 


5.0 


50.8 


4.6 


39.2 


5.2 


48.0 


4.8 


39.6 


5.0 


44.0 


4.9 



Table 5.5: Molar mass averages from GPC for 2-phenylethyl methacrylate-nadic methyl 

anhydri de-maleic anhydride copolymers 

Polymer 

Poly(PMA-co-7.5%NMA-co-22.5%MA) 

Poly (PMA-co-7.5%NMA-co-22.5%MA) Repeat 

Poly (PMA-co- 1 5%NMA-co- 1 5%MA) 

Poly (PMA-co- 15%NMA-co-15%MA) Repeat 

Poly (PMA-co-22.5%NMA-co-7.5%MA) 

Poly (PMA-co-22.5%NMA-co-7.5%MA) Repeat 



The spectrum of PMA (Figure 5.9) contains six major peaks associated with the 
different chemical environments of protons Table 5.6. There is a small peak located at 
3.5 ppm. This peak is due to contamination in one of the vials of deuterated DMF 
solvent used to make samples. The contaminant does not show up in all the systems 
because it was only in one of the vials. Figure 5.14 contains the spectra of the 
contaminated solvent. 



172 



Table 5.6: Summary of the 1H NMR Spectra from phenylethyl met hacrylate 

Chemical Group Shift (ppm) 

Aromatic-H 7.2 Singlet 

C=CH2 5.1,5.7 Singlet 

COO-CH2-C 4.4 Triplet 

Aromatic-Ctb 3.0 Triplet 

CH 2 =C-CH3 _ 1.9 Singlet 

Contaminant in d-DMF 3.5 Singlet 



The spectrum of nadic methyl anhydride (Figure 5.10) has three major regions where 
peaks occur. The most identifiable are the peaks located at 5.8 ppm that are associated 
with protons in vinyl bonds. This peak is split into a doublet by the single proton on the 
adjacent carbon. The second group of peaks is located between 3 and 4 ppm. The 
doublet peak with the largest integration is located at 3.2 ppm. This peak is due to 
protons located on carbon atoms adjacent to carbonyl groups in the anhydride linkages. 
The third group of peaks is associated with the carbons in the aliphatic linkages and the 
methyl group adjacent to the vinyl bonds. Exact peak assignments are difficult due to the 
large amount of spin couplings and complex chemical environments. There are also 
some notable peaks in the 3 to 4 ppm range. These peaks along with the low integration 
values of the vinyl proton indicate that there is some contamination in the NMA 
monomer. The normalized integration values indicate that the vinyl functionality is 
present at 38 percent of what would be expected based on the structure of NMA. The 
chemical structure of maleic anhydride indicates that there should only be a single proton 
environment (Figure 5.1 1). There are 4 peaks in maleic anhydride spectra. The peak that 
occurs at 6.4 is due to the protons in the vinyl group. The other three peaks are due to 
DMF. 



173 



tblt H '1. 9.0. 
Pull* Scitutnct: 



2.92, 2.7S «rt (>•»*' <*"i tO Mr 



jlJ 



— 1 1 1 1 1 ] r— 

8 7 



^JL. 



4r— h 



L JaJvL vli 



-i — ■ — r- 

s 



ppa 



Figure 5.9: The proton NMR spectra of 2-phenylethyl methacrylate in deuterated DMF 
referenced to TMS 



174 






The peaks in the spectrum of poly (PMA-co-40%NMA) are broader than those of 
the corresponding monomers (Figure 5.12). The broadening is due to the reduced 
mobility of the chemical groups in the polymer structure. The peaks previously identified 
in the PMA spectra at 7.3, 4.2, and 3.0 are present. The peak at 3.5 is due to 
contamination in DMF solvent. The peaks at 1.8 and 0.8 ppm are due to the protons 
located along the aliphatic backbone and the a-methyl group, respectively. The a-methyl 
peak was previously located at 1.9 ppm; however, reaction of the vinyl bond changed the 
chemical environment and resulted in an upfield shift. The values of the integration, 
which normalize to 2.86, confirm this assignment. There are no peaks present in the 5.8 
ppm range confirming that NMA incorporates through a reaction of the vinyl bonds; 
however, there are no resolved peaks that allow quantification of concentration of NMA. 
This might be expected based on the lower percentages of incorporation, i.e. 2 wt.%, 
which is below the detection level. 

The spectrum of the poly (PMA-co-7.5%NMA-co-22.5%MA) (Figure 5.13) is 
very similar to the spectrum from poly (PMA-co-40%NMA). The only notable 
difference is the magnitude of the peak at 3.5 ppm. There are no vinyl proton peaks (6.4 
ppm), which indicates that the anhydrides are incorporating through a reaction of the 
vinyl bonds. The peak 3.5 ppm is notably strong and may be due to protons of maleic 
anhydride that have been incorporated into the polymer backbone; however, proton NMR 
spectrum of poly (norbornene-co-MA) report no peaks at a chemical shift greater than 
3.15 [126]. As mentioned earlier, the peak at 3.5 is due to contamination in the DMF 
solvent (Figure 5.14). Although there is no single peak that identifies NMA or maleic 



175 



Ttil* IE *2. «.«*, 1-«, *nd US *rc ftCAks du* to £WF 

Pulse Sequence: Si put 



J^ 








-1 — I — 1 — I — I — I — I- 



-«.!*• l.tl 



pp.rn 



Figure 5.10: The proton NMR spectra of nadic methyl anhydride in deuterated DMF 
referenced to TMS 



176 



HI standard parameters 
Pule* Sequence: s2pu 1 






JL 

__uUi_> VJB-. — 



— | — i — i — i — i — r 
10 9 



LJ' 



-A_A_ 



—\ 1 1 1 1 1 1 1 1 1 1 r- 

7 6 5 



"" — I — ■" 
2 



ppn 



0.25 2.17 

2.18 -0.00 



Figure 5.11: The proton NMR spectra of maleic anhydride in DMF referenced to TMS 



177 

anhydride in these NMR spectra, the possibility of maleic anhydride being incorporated 
into the polymer backbone cannot be ruled out based on this data. The peaks 
representing NMA and MA that has been incorporated into poly (PMA) are not 
resolvable from the peaks due to poly (PMA). The possibility that NMA and MA are 
incorporating through some mechanism other than polymerization through the vinyl bond 
can be ruled out using this data. 

Conclusions 
FTIR and DSC data indicate that anhydrides are incorporating into the polymer 
backbone. NMR data confirms that incorporation is through a polymerization of the vinyl 
group. Nadic methyl anhydride (NMA) does not incorporate at levels higher than 2% 
irrespective of its feed composition in the monomer. A higher concentration of MA in 
the feed results in larger amounts of anhydrides in the polymers; however, the exact ratio 
of NMA to MA in the polymer was not determined. There also does not appear to be a 
synergistic or copolymerization effect when MA and NMA are present together in the 
monomer feed. The incorporation efficiency is less than 30% of the feed concentration 
for all systems studied in this work. This indicates that the maleic anhydride and nadic 
methyl anhydride may not be good candidates for incorporating anhydride moieties into 
methacrylate-based materials. 



tmi l| M. H.t>n 1. It, «id 1.7S *re »u«s Ova to DHF 

Pule* S»*U*nc«; tJpi/1 



178 




— i — ' — »- 

8 



1.1 - 

UULJU 




-i — i — i — i — i — j- 

3 Z 



-i 1 ' "" 

1 



ppm 



Figure 5.12: The proton NMR spectra of poly (PMA-co-40%NMA) in deuterated DMF 
referenced to TMS 



179 



TMs is IV. "*4* «»**! 1-03, M, ftM 2.7 «r# pMks «u* to IM DHF 
Putia S«quaftc«: b$pul 



3 



Ul 



1 1 1 1 T"- 

I 




Figure 5.13: The proton NMR spectra of poly (PMA-co-7.5%NMA-co-22.5%MA) in 
deuterated DMF referenced to TMS 




u 






1.00 



Figure 5.14: The proton NMR spectra of deuterated DMF referenced to TMS 



CHAPTER 6 
CLOSING REMARKS 

In Chapter 3, the monofunctional methacrylates cyclohexyl methacrylate 
(CHMA), phenyloxyethyl methacrylate (PEMA), and tert-butylcyclohexyl methacrylate 
(t-BCHMA) were introduced as alternate diluent systems for Bis-MEPP-type monomers 
for dental applications. The monofuctional methacrylates reduced polymerization 
shrinkage by 10 to 20 percent when compared to Bis-MEPP/TEGDMA systems. The 
reduction in shrinkage is due to the higher specific volume per methacrylate group of 
monomethacrylates compared to TEGDMA. CHMA and PEMA reduced the viscosity of 
Bis-MEPP more effectively than TEGDMA, while t-BCHMA did not. The Bis- 
MEPP/CHMA and Bis-MEPP/PEMA systems had similar moduli, 2.8 ± 0.1 and 2.7 ± 0.1 
GPa for the Bis-MEPP/CHMA and Bis-MEPP/PEMA systems in wet state at 37°C, 
respectively, to the Bis-MEPP/TEGDMA systems (2.6 ± 0.1 GPa). Quantitative 
fractography was applied to measure fracture toughness. To the author's knowledge, this 
is the first time this technique has been used to measure fracture toughness in dental 
polymers. The fracture toughness of the Bis-MEPP/PEMA and Bis-MEPP/CHMA 
systems was an average of 0.2 MPa.m ' 5 greater than the Bis-MEPP/TEGDMA systems. 
The modulus and fracture toughness were lower in the Bis-MEPP/t-BCHMA systems 
than in the Bis-MEPP/TEDGMA systems. The reduced properties might in part be due 
to the high glass transition temperatures and viscosities of the Bis-MEPP/t-BCHMA 
monomer systems. A non-reactive contaminant was also present that might result in the 
reduction of the mechanical properties of these systems. Multicomponent diluent 

181 



182 

systems based on variable molar mass analogs of ethylene glycol dimethacrylate were 
also explored as alternate diluent systems. The modulus and strength properties of the 
multicomponent systems did not vary greatly from those of the Bis-MEPP/TEGDMA 
polymers. The similarity of the properties of all the systems examined in Chapter 3 is 
attributed to the heterogeneous network structure that results from the incomplete cure as 
well as the mechanisms of network formation. The dimethacrylate polymers are 
composed of regions of high and low cross-link density. This structure is indicated by 
the DMS data presented later in Chapter 3 as well as the work presented in Chapter 4. 

Monomer systems based on Bis-MEPP/CHMA Bis-MEPP/PEMA and Bis- 
MEPP/t-BCHMA were formulated to produce polymers with the same glass transition 
temperature. Formulating the polymers to the same Tg allows the isolation of the effect 
of initial viscosity on vitrification and heterogeneity of these systems. Initial viscosity 
was shown to be an indicator of heterogeneity of the polymers formed rather than a 
primary factor. The details of network formation, phase segregation, and isomeric forms 
of the monomers have strong effects on the heterogeneity of the polymers formed. 

The heterogeneity of dimethacrylate polymers was elucidated by xenon NMR, 
Tapping Mode™AFM measurements using phase imaging, and dynamic mechanical 
spectroscopy. Xenon spectra of a Bis-MEPP(XDT) polymer cured for 5 minutes with 
UV light showed a peak indicative of a polymer that is not fully glassy. The width and 
intensity of the peak suggest that a portion of the polymer has a pore structure that is not 
representative of a fully glassy polymer. Pulsed Field Gradient 129 xenon NMR was used 
to measure the change in domain sizes of the sol phase with cure in the Bis-MEPP(XDT) 
systems. It is thought that heterogeneity has its roots early in the cure cycle. Measuring 



183 

how the sol phase changes with cure would lead to a better understanding of the 
mechanisms that yield a heterogeneous structure. Unfortunately, measurements could 
only be made at low conversions. Equipment in which stronger field gradients can be 
applied would allow the measurement to be pushed further into the cure cycle. 
Variations in the cross-link density of dimethacrylate type polymers on the order of 25- 
150 nm was demonstrated using Tapping Mode™AFM measurements using phase 
imaging. Samples were etched in acetone and post cured to help make the assignment of 
different regions of phase contrast to the low and high cross-link density phases. In these 
systems the high cross-link density phases have a greater phase shift than the low cross- 
link density phase. The presence of the low temperature shoulder on the main transition 
in the tan 5 plots of the polymers tested in Chapter 3 is also indicative of a bimodal 
distribution of cross-link density. 

Characterization of the size and distribution of the low and high cross-link density 
phases is important because the low cross-link density phases of these polymers dictate 
their mechanical properties. Further quantification of the size of these structures with 
image analysis techniques might allow correlations to be drawn between their distribution 
and mechanical properties. These correlations may ultimately allow the tailoring of 
monomer selection and cure cycles to improve mechanical properties. 

The efficiency of the incorporation of nadic methyl anhydride (NMA) and maleic 
(MA) anhydride into 2-phenyl ethyl methacrylate polymers was determined using FTIR, 
GPC, DSC, and *H NMR. This study was performed to glean information about how 
NMA and MA incorporate into aromatic dimethacrylate polymers such as Bis-GMA and 
Bis-MEPP. The results indicate that nadic methyl anhydride does not incorporate at 



184 

amounts greater than 2 mol.% regardless of the feed concentrations. The amount of 
anhydride present in the polymers increased as the concentration of maleic anhydride was 
increased in the monomer feed. The exact ratios of the MA and NMA were not 
determined. The incorporation efficiency of all the anhydrides in this study was less than 
30 mol.% of the feed concentration, thus indicating that these anhydrides may not be 
good candidates for comonomers in methacrylate based dental polymers. The 
incorporation of anhydride groups into methacrylate dental polymers might be better 
facilitated by monomers that have a methacrylate functionality or a functionality that is 
known to polymerize with methacrylates. 






LIST OF REFERENCES 



1 . Holter, D., et al. Branched Bismethacrylates Based on Bis-GMA - A Systematic 
Route to Low Shrinkage Composites, in Proceedings of the 1997 Las Vegas ACS 
Meeting. 1997. Las Vegas, NV, USA: Division of Polymer Chemistry, American 
Chemical Society. 

2. Sankarapandian, M., et al. Synthesis of New Dental Composite Matrix 
Dimethacrylates. in Proceedings of the 1997 Las Vegas ACS Meeting. 1997. Las 
Vegas, NV, USA: Division of Polymer Chemistry, American Chemical Society. 

3. Ferracane, J.L. and E.H. Greener, The Effect of Resin Formulation on the Degree 
of Conversion and Mechanical Properties of Dental Restorative Resins. Journal 
of Biomedical Materials Research, 1986. 20: p. 121-131. 

4. Shobha, H.K., et al., Structure property relationship among novel dental 
composite matrix resins. Journal of Materials Science: Materials in Medicine, 
1997. 8: p. 385-389. 

5. Kalachandra, S., Glassy Polymers as Dental Materials, in Encylopedia Handbook 
of Biomaterials and Bioengineering, D.J.T. D. L. Wise, D. E. Antobelli, M. J. 
Yaszemski, J. D. Gresser, and E. R. Schwartz, Editor. 1995, Marcel Dekker, Inc.: 
New York. p. 1545-1556. 

6. Liu, C.-F., CM. Collard, and CD. Armeniades, Constant -volume polymerization 
of composites by addition of ammonia-modified montmorillonite. American 
Journal of Dentistry, 1990. 3(2)( April): p. 44-50. 

7. Stansbury, J. W., Synthesis and Evaluation of New Oxaspiro Monomers for 
Double Ring-opening Polymerization. Journal of Dental Research, 1991. 
71(7)(July): p. 1408-1412. 

8. Peutzfeldt, A. and E. Asmussen, Influence ofCarboxylic Anhydrides on Selected 
Mechanical Properties of Heat-cured Resin Composites. Journal of Dental 
Research, 1991. 70(12)(December): p. 1537-1541. 

9. Simon, G.P., et al., Nature of Residual Unsaturation during Cure of 

Dimethacrylates Examined by CPPEMAS ' " C and Simulation Using a Kinetic 
Gelation Model. Macromolecules, 1989. 22: p. 3555-3561. 



185 



186 



10. Chiu, Y.Y. and L.J. Lee, Microgel Formation in the Free Radical Crosslinking 
Polymerization of Ethylene Glycol Dimethacrylate (EDGMA). II. Simulation. 
Journal of Polymer Science: Part A: Polymer Chemistry, 1995. 33: p. 269-283. 

1 1 . Chiu, Y.Y. and L.J. Lee, Microgel Formation in the Free Radical Crosslinking 
Polymerization of Ethylene Glycol Dimethacrylate (EGDMA). I. Experimental. 
Journal of Polymer Science: Part A: Polymer Chemistry, 1995. 33: p. 257-267. 

12. Watts, D.C., Filling the Gaps. Chemistry in Britain, 1992. 28(3): p. 235 - 238. 

13. Peutzfeldt, A., Resin composites in dentistry: the monomer systems. European 
Journal of Oral Sciences, 1997. 105: p. 97-116. 

14. Jones, D.W., Synthesis and evaluation of ceramic, polymer, and composite 
biomaterials, in Encylopedia Handbook of Biomaterials and Bioengineering, 
D.J.T. D. L. Wise, D. E. Antobelli, M. J. Yaszemski, J. D. Gresser, and E. R. 
Schwartz, Editor. 1995, Marcel Dekker, Inc.: New York. p. 187 - 202. 

15. Prati, C, et al., Dental Composite Resin Porosity and Effect on Water Absorption. 
Boll. Soc. It. Biol. Sper., 1991. LXVII(4). 

16. Phillips, R.W., Elements of Dental Materials. Fourth ed. 1984, Philadelphia: 
Saunders. 

17. Yamaguchi, R., J.M. Powers, and J.P. Dennison, Thermal Expansion of visible- 
light-cured composite resins. Operative Dentistry, 1989. 14: p. 64-67. 

18. Taylor, D.F., et al., Relationship between fdler and matrix resin characteristics 
and the properties of uncured composite pastes. Biomaterials, 1998. 19: p. 197- 
204. 

19. Shobha, H.K., et al., Effect of dilution on the kinetics of cross-linking thermal 
polymerization of dental composite matrix resins. Journal of Materials Science: 
Materials in Medicine, 1997. 8: p. 583-586. 

20. Venhoven, B.A.M., et al., Infuence of fdler parameters on the mechanical 
coherence of dental restorative resin composites. Biomaterials, 1996. 17(7): p. 
735-740. 

21. Jones, D.W. and A.S. Rizkalla, Characterization of Experimental Composite 
Biomaterials. Journal of Biomedical Materials Research, 1996. 33: p. 89-100. 

22. Willems, G.P., et al., A classification of dental composites according to their 
morphological and mechanical characteristics. Dental Materials, 1992. 
8(September): p. 310-319. 





187 


23. 


Craig, R.G., W. J. O'Brien, and J. M. Powers, Dental Materials .Properties and 
Manipulation. 6th ed. 1996, St. Louis: Mosby. 55 - 75. 


24. 


Chernova, Z.D., et al., Densely Crosslinked Polymer Compositions Based on 
Various Types ofDimethacrylates. Polymer Science U.S.S.R., 1981. 23(10): p. 
2438-2445. 


25. 


Kannurpatti, A.R. and C.N. Bowman, Structural Evolution of Dimethacrylate 
Networks Studied by Dielectric Spectoscopy. Macromolecules, 1998. 31(10): p. 
3311-3316. 


26. 


Kannurpatti, A.R., J.W. Anseth, and C.N. Bowman, A study of the evolution of 
mechanical properties and structural hertogeneity of polymer networks formed by 
photopolymerizations of multifunctional (meth)acrylates. Polymer, 1998. 39(12): 
p. 2507-2513. 


27. 


Kannurpatti, A.R., et al., Use of "Living" Radical Polymerizations to Study the 
Structural Evolution and Properites of Highly Crosslinked Polymer Networks. 
Journal of Polymer Science: Part B: Polymer Physics, 1997. 35: p. 2297-2307. 


28. 


Simon, G.P., P.E.M. Allen, and D.R.G. Williams, Properties of dimethacrylate 
copolymers of varying crosslink density. Polymer, 1991. 32(14): p. 2577-2587. 


29. 


Ferracane, J.L. and J.R. Condon, Rate ofElution ofLeachable Components from 
composite. Dental Materials, 1990. 6: p. 282-287. 


30. 


Simon, G.P., P.E.M. Allen, and D.R.G. Williams, Examination of Dimethacrylate 
Inhomogeneity by Solvent Swelling Tehcniques. Polymer Engineering and 
Science, 1991. 31(20): p. 1483-1492. 


31. 


Spahl, W., H. Budzikiewicz, and W. Geurtsen, Determination ofleachable 
components from four commercial dental composites by gas and liquid 
chromatography /mass spectrometry. Journal of Dentistry, 1998. 26(2): p. 137- 
145. 


32. 


Anseth, K.S., M.D. Rothenberg, and C.N. Bowman, A Photochromic Technique 
To Study Polymer Network Volume Distributions and Microstructure during 
Photopolymerizations. Macromolecules, 1994. 27(10): p. 2890-2892. 


33. 


Lange, J., R. Ekelof, and G.A. George, Indications of Micro-vitrification during 
chainwise cross-linking poymerisation. Polymer, 1999. 40: p. 3595-3598. 


34. 


Golikov, I. V., et al., Formation of the Microstructure and Strength Properties of 
Oligoester Acrylate Polymers. Polymer Science U.S.S.R., 1979. 21(8): p. 2013- 
2022. 



188 



35. Baidin, I.S., et al., Effect of the Structure of a Three-Demensional Network of the 
Polymer Triethylene Glycol Dimethacrylate on the Kinetics of its Radical 
Polymerization. Polymer Science USSR, 1989. 31(7): p. 1525-1529. 

36. Erath, E.H. and R.A. Spur, Occurrence of Globular Formation in Thermosetting 
Resins. Journal of Polymer Science, 1959. 35: p. 391-399. 

37. Kannurpatti, A.R., et al. Polymerization Behavior and Properties of Networks 
formed by Dimethacrylate Dental Resins, in Proceedings of the 1997 Las Vegas 
ACS Meeting. 1997. Las Vegas, NV, USA: Division of Polymer Chemistry, 
American Chemical Society. 

38. Sandner, B., et al., Synthesis ofBISGMA derivatives, properties of their polymers 
and composites. Journal of Materials Science: Materials in Medicine, 1997. 8: p. 
39-44. 

39. Kalachandra, S., et al., Polymer Materials for Composite Matrices in Biological 
Environment. Polymer, 1993. 34(4): p. 778-782. 

40. Stansbury, J.W., M.C. Kyung, and J.M. Antonucci. Consideration in the 
Development of Semi-fluorinated Methacrylate Dental Resins and Composites, in 
Proceedings of the 1997 Las Vegas ACS Meeting. 1997. Las Vegas, NV, USA: 
Division of Polymer Chemistry, American Chemical Society. 

41. Kawaguchi, M., T. Fukushima, and T. Horibe, Effect of Monomer Structure on 
the Mechanical Properties of Light-cured Unfilled Resins. Dental Materials 
Journal, 1988. 7(2): p. 174-181. 

42. Chowdhury, N.A., et al., Dental application of binary urethane monomer 
mixtures: strengthened resin matrix. Journal of Materials Science: Materials in 
Medicine, 1997. 8: p. 149-155. 

43. Mitra, S.B. Dental Composites Prepared from Resin Matrices Containing 
Ethylenically Unsaturated Carbamoyl Isocyanurates. in Proceedings of the 1997 
Las Vegas ACS Meeting. 1997. Las Vegas, NV, USA: Division of Polymer 
Chemistry, American Chemical Society. 

44. Culbertson, B.M., Q. Wan, and Y. Tong. Synthesis, Characterization and 
Evaluation of New BP A Based Multi-Methacrylates for Dental Composites, in 
Proceedings of the 1997 Las Vegas ACS Meeting. 1997. Las Vegas, NV, USA: 
Division of Polymer Chemistry, American Chemical Society. 

45. Dulik, D., R. Bernier, and G.M. Brauer, Effect of Diluent Monomer on the 
Physical Properties of Bis-GMA-based Composites. Journal of Dental Research, 
1981. 60(6): p. 983-989. 



189 



46. Davy, K.W.M. and M. Braden, Study of polymeric systems based on 2,2 bis-4(2- 
hydroxy-3-methacyloyl-oxypropoxy) phenyl propane. Biomaterials, 1 99 1 . 
12(May): p. 406-410. 

47. Labella, R., et al., THFMA in dental monomer systems. Biomaterials, 1996. 17(4): 
p. 431-436. 

48. Labella, R., et al., Monomethacrylate co-monomers for dental resins. European 
Journal of Oral Sciences, 1998. 106: p. 816-824. 

49. Patel, M.P. and M. Braden, Heterocyclic Methacrylates for clinical applications I. 
Mechanical properties. Biomaterials, 1991. 12(September): p. 645-648. 

50. Cowperthwaite, G.F., J.J. Foy, and M.A. Malloy, The Nature of the Cross-linking 
Matrix Found in Dental composite Filling Materials and Sealants, in Biomedical 
and Dental Applications of Polymers, C.G. Gebelein and F.F. Koblitz, Editors. 
1981, Plenum Press: New York. p. 379. 

5 1 . Patel, M.P., M. Braden, and K.W.M. Davy, Polymerization Shrinkage of 
Methacry late Esters. Biomaterials, 1987. 8: p. 53-56. 

52. Loshaek, S. and T.G. Fox, Cross-linked Polymers. I. Factors Influencing the 
Efficiency of Cross-linking in Copolymers on Methyl Methacrylate and Glycol 
Dimethacrylates. Journal of the American Chemical Society, 1953. 75: p. 3544- 
3550. 

53. Bowen, R.L. Preparation of a Beta-Cyclodextrin Methacrylate. in Proceedings of 
the 1997 Las Vegas ACS Meeting. 1997. Las Vegas, NV: Division of Polymer 
Chemistry, American Chemical Society. 

54. Kalachandra, S., et al., Infuence of Hydrogen Bonding on Properties ofBIS-GMA 
Analogues. Journal of Materials Science: Materials in Medicine, 1997. 8: p. 283- 
286. 

55. Rawls, H.R., et al. Low Shrinkage Resins From Liquid Crystal Diacrylate 
Monomers, in Proceedings of the 1997 Las Vegas ACS Meeting. 1997. Las Vegas, 
NV, USA: Division of Polymer Chemistry, American Chemical Society. 

56. Byerley, T.J., et al., Synthesis and Polymerization of New Expanding Dental 
Monomers. Dental Materials, 1992. 8: p. 345-350. 

57. Chappelow, C.C., et al. Photocured Epoxy/SOC Matrix Resin Systems for Dental 
Composites, in Proceedings of the 1997 Las Vegas ACS Meeting. 1997. Las 
Vegas, NV, USA: Division of Polymer Chemistry, American Chemical Society. 






190 

58. Millich, F., et al., Elements of Light-Cured Epoxy-based Dental Polymer Systems. 
Journal of Dental Research, 1998. 77(4): p. 603-608. 

59. Moszner, N. and T. Volkel. Radical Ring-Opening Monomers for Dental 
Composites, in Proceedings of the 1997 Las Vegas ACS Meeting. 1997. Las 
Vegas, NV, USA: Division of Polymer Chemistry, American Chemical Society. 

60. Miyazaki, K., et al. Synthesis and Polymerization of Acrylic Monomers with 
Pendant Spiro Ortho Ester and Cyclic Carbonate Groups, in Proceedings of the 
1997 Las Vegas ACS Meeting. 1997. Las Vegas, NV, USA: Division of Polymer 
Chemistry, American Chemical Society. 

61. Yoshida, K., H. Matsumura, and M. Atsuta, Monomer Composition and Bond 
Strength of Light Cured 4-META Opaque Resin. Journal of Dental Research, 
1990. 69(3)(March): p. 849-851. 

62. Peutzfeldt, A. and E. Asmussen, Effect of temperature and duration of post-cure 
on selected mechanical properties of resin composites containing carboxylic 
anhydrides. Scandinavian Journal of Dental Research, 1992. 100: p. 296-298. 

63. Peutzfeldt, A., Effect of Temperature and Duration ofPostcure On in-Vitro Wear 
and Quantity of Remaining Double-Bonds of Resins Containing Carboxylic 
Anhydride. European Journal of Oral Sciences, 1995. 103(4): p. 259-263. 

64. Asmussen, E. and A. Peutzfeldt, The role ofmaleic anhydride in adhesive resin 
cements. European Journal of Oral Sciences, 1998. 106: p. 882-886. 

65. Ishihara, K. and N. Nakabayashi, Adhesive bone cement both to bone and 
metals:4-META in MMA with tri-n-butyl borane. Journal of Biomedical Materials 
Research, 1989. 23: p. 1475-1482. 

66. Ishihara, K., et al., Adhesive bone cement containing hydroxyapatite particle as 
bone compatible filler. Journal of Biomedical Materials Research, 1992. 26: p. 
937-945. 

67. West, J.K., et al., Cyclic anhydride ring opening reactions: Theory and 
application. Journal of Biomedical Materials Research, 1998. 41(1): p. 8-17. 

68. Zamora, M.P., Poly(Amide-Graft-Acrylate) Interfacial Compounds, in Materials 
Science and Engineering. 1997, University of Florida: Gainesville, p. 185. 

69. Bowen, R.L., E.N. Cobb, and J.E. Rapson, Adhesive Bonding of Various 
Monomers to Hard Tooth Tissues Improvement in Bond Strength to Dentin. 
Journal of Dental Research, 1982. 61(9): p. 1070-1076. 



191 



70. Bowen, R.L., et al., New Surface-active comonomer for Adhesive Bonding. 
Journal of Dental Research, 1996. 75(1): p. 606-610. 

7 1 . Kalachandra, S. and R.P. Kusy, Comparison of water sorption by methacrylate 
and dimethacrylate monomers and their corresponding polymers. Polymer, 1991. 
32(13): p. 2428-2434. 

72. Soderholm, K.-J., Water sorption in a bis(GMA)/TEGDMA resin. Journal of 
Biomedical Materials Research, 1984. 18: p. 271-279. 

73. Turner, D.T. and A.K. Abell, Water sorption of Poyl(methyl methacrylate): 2. 
Effects of crosslinks. Polymer, 1987. 28: p. 297-302. 

74. Miller, J.B., J.H. Walton, and CM. Roland, The NMR Chemical Shift ofXenon- 
129 Dissolved in Polymers. Macromolecules, 1993. 26(21): p. 5602-5610. 

75. Sefcik, M.D., et al. Effects ofSorbed Gases on the Molecular Dynamics and 
Structure of Glassy Polymers, in Polymer Preprints. 1983: American Chemical 
Society, Divsion of Polymer Chemistry. 

76. Stengle, T.R. and K.L. Williamson, Nuclear Magnetic Resonance of Xenon 
Absorbed in Solid Polymers: A Probe of the Amorphous State. Macromolecules, 
1987. 20: p. 1428-1430. 

77. Kentgens, A.P.M., et al., Line-Broading Effects for *2*Xe Absorbed in the 
Amorphous State of Solid Polymers. Macromolecules, 1991. 24: p. 3712-3714. 

78. Kennedy, G.J., *29Xe NMR as a probe of the effect of crosslinking on the 
amorphous phase structure of solid polymer. Polymer Bulletin, 1990. 23: p. 605- 
608. 

79. Walton, J.H., A Review of^^Xe NMR as a Probe of Polymer Morphology. 
Polymers and Polymer Composites, 1994. 2(1): p. 35-41. 

80. Chu, P.J., et al. . in International Symposium on Solid State NMR Spectroscopy of 
Polymers. 1993. Keystone, CO. 

81. Morgan, D.R., E.O. Stejskal, and A.L. Andrady, ***Xe NMR Investigation of the 
Free Volume in Dendritic and Cross-Linked Polymers. Macromolecules, 1999. 
32(6): p. 1897-1903. 

82. Tomaselli, M., et al., Probing Microheterogeneity in Polymer Systems Via Two- 

Dimensional *2* Xenon NMR Spy Detection. A Heterogeneous Model Blend 
System. Chemical Physics Letters, 1993. 205(2,3): p. 145-152. 



192 



83. Mansfeld, M. and W.S. Veeman, } H- 129 Xe double resonance NMR in a polymer 
blend. Chemical Physics Letters, 1994. 222( 27 May): p. 422-424. 

84. Walton, J.H., J.B. Miller, and CM. Roland, n9 Xe NMR as a Probe of Polymer 
Blends. Journal of Applied Polymer : Part B: Polymer Physics, 1992. 30: p. 527- 
532. 

85. Mirabella, F.M.J, and D.C. McFaddin, 129 Xe NMR Spectroscopic 
Characterization of Multiphase Polypropylene Copolymers. Polymer, 1996. 
37(6): p. 931-938. 

86. Junker, F. and W.S. Veeman, Xenon Self-Diffusion on Organic Polymers by 
Pulsed Field Gradient NMR Spectroscopy. Macromolecules, 1998. 31: p. 7010- 
7013. 

87. Schantz, S. and W.S. Veeman, NanoHeterogeneity in PEO/PMMA Blends 
Probed. Journal of Polymer Science, Part B: Polymer Physics, 1997. 35: p. 2681- 
2688. 

88. Walton, J.H., et al., Phase Transitions in Polymer Blends via *™Xe NMR 
Spectroscopy. Macromolecules, 1993. 26(15): p. 4052-4054. 

89. Brownstein, S.K., J.E.L. Roovers, and D.J. Worsfold, '^Xe Line widths in Block 
Copolymers. Magnetic Resonance in Chemistry, 1988. 26: p. 392-393. 

90. Stejskal, E.O. and J.E. Tanner, Spin Diffusion Measurements: Spin Echoes in the 
Presence of a Time-Dependent Field Gradient. Journal of Chemical Physics, 
1965. 42(1): p. 288-292. 

91. Whiting, R. and P.H. Jacobsen, Dynamic Mechanical Properties of Resin-based 
Filling Materials. Journal of Dental Research, 1980. 59(1)( January): p. 55-60. 

92. Wilson, T.W. and D.T. Turner, Characterization of Polydimethacrylates and 
Their Composites by Dynamic Mechanical Analysis. Journal of Dental Research, 
1987. 66(5): p. 1032-1035. 

93. Clarke, R.L., Dynamic mechanical thermal analysis of dental polymers II. Bis- 
phenol A-r elated resins. Biomaterials, 1989. lO(October): p. 549-552. 

94. Wilson, T.W., Multiple Transitions in a Dimethacrylate Network. Journal of 
Applied Polymer Science, 1990. 40: p. 1 195-1208. 

95. Allen, P.E.M., et al., Dynamic-Mechanical Properties and Cross-Polarized, 
Proton-Enhanced, Magic Angle Spinning " C NMR Time Constants of 



193 



Poly [oligo(ethylene glycol) dimethacrylates] '. Macromolecules, 1989. 22: p. 809- 
816. 

96. McCrum, N.G., B.E. Read, and G. Williams, Anelastic and Dielectric Effects in 
Polymeric Solids. 1991, New York: Dover Publications Inc. 617. 

97. Nichetti, D. and I. Manas-Zloczower, Viscosity model for polydisperse polymer 
melts. Journal of Rheology, 1998. 42(4): p. 951-969. 

98. Plangsangmas, L., Determination of Fracture Toughness ofEpoxy Using 
Fractography. Journal of Applied Polymer Science, 1998. 72: p. 257-268. 

99. Mehlem, J.J., The Influence ofNadic Methyl Anydride on Mechanical and 
Thermomechanical Properties of Dental Resins and Composites, in Materials 
Science and Engineering. 1999, University of Florida: Gainesville, p. 1 15. 

1 00. Nomoto, R. and T. Hirasawa, Residual Monomer and Pendant Methacryloyl 
Group in Light-cured Composite Resins. Dental Materials Journal, 1992. 11(2): p. 
177-188. 

101. Allcock, H.R. and F.W. Lampe, Contemporary Polymer Chemistry. Second ed. 
1990, Englewood Cliffs: Prentice-Hall, Inc. 624. 

102. Ferracane, J.L., Fourier Transform Infrared Analysis of Degree of Polymerization 
in Unfilled Resins-Methods Comparison. Journal of Dental Research, 1984. 
63(8)(August): p. 1093-1095. 

103. Asmussen, E., Factors affecting the quantity of remaining double bonds in 
restorative resin polymers. Scandinavian Journal of Dental Research, 1982. 90: p. 
490-496. 

104. Ruyter, E.I. and S.A. Svendsen, Remaining methacrylate groups in composite 
restorative materials. Acta Odontol. Scand., 1977. 36: p. 75-82. 

105. Rueggeberg, F.A., D.T. Hashinger, and C.W. Fairhurst, Calibration ofFTIR 
conversion analysis of contemporary dental resin composites. Dental Materials, 
1990. 6(October): p. 241-249. 

106. Bastioli, C, G. Romano, and C. Migliaresi, Water sorption and mechanical 
properties of dental composites. Biomaterials, 1990. ll(April): p. 219-223. 

107. Soderholm, K.-J.M. and M.J. Roberts, Influence of Water Exposure on the Tensile 
Strength of Composites. Journal of Dental Research, 1989. 69(12): p. 1812-1816. 



194 

1 08. Kloosterboer, J.G. and G.F.C.M. Lijten, Photopolymers exhibiting a large 
difference between glass transition and curing temperatures. Polymer, 1990. 
31(January): p. 95-101. 

109. Floudas, G., G. Fytas, and E.W. Fischer, Relaxation Porcesses in a 
Poly(cyclohexyl methacrylate) /Additive System As Studied by Photon Correlation, 
Dielectric Relaxation, and Mechanical Relaxation Spectroscopy. 
Macromolecules, 1991. 24(8): p. 1955-1961. 

1 10. Sperling, L.H., Introduction to Physical Polymer Science. Second ed. 1992, New 
York: John Wiley & Sons. 594. 

111. Kannurpatti, A.R., et al., Kinetic and Mechanistic Studies oflniferter 
Photopolymerizations. Macromolecules, 1996. 29(23): p. 7310-7315. 

1 12. Simpson, J.H. and H.Y. Carr, Diffusion and Nuclear Spin Relaxation in Water. 
The Physical Review, 1958. 111(5): p. 1201-1202. 

113. Bar, G, et al., Examination ofButadiene/Styrene-co-butadience Rubber Blends by 
Tapping Mode Atomic Force Microscopy. Tip-Sample Energy Dissipation in 
Tapping Mode Atomic Force Microscopy study of Elastomers. Langmuir, 2000. 
16(13): p. 5702-5711. 

1 14. Bar, G, et al., Factors Affecting the Height and Phase Images in Tapping Mode 
Atomic Force Microscopy. Study of Phase-Separated Polymer Blends of 
Poly(Ethene-co-Styrene) and Poly(2,6-dimethyl-l ,4-phenylene oxide). Langmuir, 
1997. 13(14): p. 3807-3812. 

115. Magonov, S.N., V. Elings, and M.-H. Whangbo, Phase imaging and stiffness in 
tapping-mode atomic force microscopy. Surface Science, 1997. 375: p. L385- 
L391. 

1 16. Thomann, Y., et al., Investigation of Morphologies of One- and Two-Phase 
Belnds oflsotactic Poly(propene) with Random Poly(ethene-co-I-butene). 
Macromolecules, 1998. 31(16): p. 5441-5449. 

1 17. Kajiyama, T., et al., Imaging of Dynamic Viscoelastic Properties of a Phase- 
Separated Polymer Surface by Forced Oscillation Atomic Force Microscopy. 
Macromolecules, 1994. 227(26): p. 7932-7934. 

1 1 8. Leclere, P., et al., Microdomain Morphology Analysis of Block Copolymers by 
Atomic Force Microscoy with Phase Detection Imaging. Langmuir, 1996. 12(18): 
p. 4317-4320. 



• 



195 

119. Bar, G., et al, Examination of the relationship between phase shift and energy 
dissipation in tapping mode atomic force microscopy by frequency-sweep and 
force probe measurements. Surface Science, 2000. 444: p. LI 1-L16. 

120. Bar, G., et al., Improtance of the indentation depth in tapping-mode atomic force 
microscopy study of compliant materials. Applied Physics Letters, 1999. 75(26): 
p. 4198-4200. 

121. Rosen, S.L., Fundamental Principles of Polymeric Materials. 1 993, New York: 
John Wiley and Sons. 420. 

1 22. Stejskal, E.O. and J.D. Memory, High Resolution NMR in the Solid State. 1 994, 
New York: Oxford University Press. 189. 

123. Stansbury, J.W. and J.M. Antonucci, Evaluation of Methylene lactone monomers 
in dental resins. Dental Materials, 1992. 8: p. 270-273. 

124. Zamora, M., K.B. Wagener, and A.B. Brennan. Copolymer composites designed 
to offset polymerization shrinkage in dental composites: I. maleic anhydride. 
1994: Division of Polymer Chemistry, American Chemical Society. 

125. Dorkenoo, K.D., P.H. Pfromm, and M.E. Rezac, Gas Transport Porperties of a 
Series of High Tg Polynorbornenes with Aliphatic Pendant Groups. Journal of 
Polymer Science: Part B: Polymer Physics, 1998. 36: p. 797-803. 

126. Houlihan, F.M., et al., Synthesis of Cycloolefin-Maleic Anhydride Alternating 
Copolymer for 193nm Imaging. Macromolecules, 1997. 30(21): p. 6517-6524. 



BIOGRAPHICAL SKETCH 

Jeremy Mehlem was born to John and Sheila Mehlem on the afternoon of October 
30, 1972, in Phoenix, Arizona. He grew up in Arizona, with an occasional summer 
escape to New Jersey to visit relatives. During high school Mehlem was involved in 
extracurricular activities such as swimming and tennis. While in high school, he also 
developed interests in mountain biking and windsurfing, despite the desert's arid climate. 

After graduating from high school in 1990, Mehlem attended the University of 
Arizona, in Tucson, graduating with a Bachelor of Science degree in materials science 
and engineering in December of 1 994. 

After travel in Mexico and the Pacific Northwest, Mehlem began graduate studies 
at the University of Florida in 1995. Unaware that Mehlem's father was a dentist, Dr. 
Anthony Brennan introduced Mehlem to research in dental materials, the subject of his 
master's thesis and this dissertation. He married Ms. Kacy Gapinski in November of 
1999. Upon graduating he will begin work with Michelin in Greenville, South Carolina. 



196 



I certify that I have read this study and that in my opinion it conforms to 
acceptable standards of scholarly presentation and is fully adequate, in scope and quality, 
as a dissertation for the degree of Doctor of Philosophy. 



Dr. Anthony Bifennan, Chairman 
Associate Professor of Materials Science 
and Engineering 



I certify that I have read this study and that in my opinion it conforms to 
acceptable standards of scholarly presentation and is fully adequate, in scope and quality, 
as a dissertation for the degree of Doctor of Philosophy 



'^JKcXa 



Dr. Christopher Batich 
Professor of Materials Science and 
Engineering 



I certify that I have read this study and that in my opinion it conforms to 
acceptable standards of scholarly presentation and is fully adequate, in scope and quality, 
as a dissertation for the degree of Doctor of Philosophy. 




Dr. Elliot D^ 
Assistant Professor of Materials Science 
and Engineering 

I certify that I have read this study and that in my opinion it conforms to 
acceptable standards of scholarly presentation and is fully adequate, in scope and quality, 
as a dissertation for the degree of Doctor of Philosophy. 

Dr. Ronald Baney £/ 
Associate Engineer of Materials Science 
and Engineering 



I certify that I have read this study and that in my opinion it conforms to 
acceptable standards of scholarly presentation and is fully adequate, in scope and quality, 
as a dissertation for the degree of Doctor of Philosophy. 



Dr.' Kenneth Wagener / 
Professor of Chemistry 

This dissertation was submitted to the Graduate Faculty of the College of 
Engineering and to the Graduate School and was accepted as partial fulfillment of the 
requirements for the degree of Doctor of Philosophy. 



May 2001 



M. J. Ohanian 

Dean, College of Engineering 



Winfred M. Phillips 
Dean, Graduate School 







MMfl? 



MMp